CHAPTER 4 RESEARCH RESULTS AND FINDINGS

University of Pretoria – C Olckers (2011)

CHAPTER 4 – RESULTS AND FINDINGS 62

CHAPTER 4

RESEARCH RESULTS AND FINDINGS

If we knew what it was we were doing, it would not be called research, would it?

- Albert Einstein

In this chapter...

��

��

��

� ��

��

��

��

��

��

��

��

��

��

#��

! ��

"��

� ��

��

��

$��

"��

��

��

��

��


CHAPTER 4 – RESULTS AND FINDINGS Page 163

4.1 INTRODUCTION

In this chapter the research methodology, strategy and research methods used to

develop a multi-dimensional measure of psychological ownership of employees within

the South African context will be documented and explained.

As explained in Chapter 3, in this study a measure of psychological ownership was

developed by following a combination of steps as suggested by DeVellis (2003), Hinkin

(1998), and Spector (1992). The various steps in the scale development process that

were followed are outlined in Figure 4.1 and are further described.

Figure 4.1: Steps in the scale development process

��3

��2

��4

��5

��6

��7

��0

��8

9��+��

:��! (��

9��! ��8��! ��

��! ��

$��! ��

��! ��! ��(! ��! (��

;��+��(��- ��#��(��

9��! ��+��! ��+�2 ��

Source: Adapted from DeVellis (2003), Hinkin (1998), and Spector (1992)



4.2 STEPS TO BE FOLLOWED IN SCALE DEVELOPMENT

4.2.1 Step 1: Defining the construct

Building on the five recognised dimensions of psychological ownership: self-efficacy,

self-identity, having a place (belonging), accountability, and territoriality of Avey and

colleagues (2009), and after a comprehensive review of the literature, the concepts of

responsibility and autonomy were posited as additional aspects of psychological

ownership. These concepts have been clearly defined and described in Chapter 1 and

in the literature review in Chapter 2.

The definition of each concept in the organisational context is summarised in Table 4.1.

Table 4.1: Definitions of key concepts in the organisational context

Concept Definition

Psychological ownership A state in which individuals feel as though the target of ownership

(material or immaterial in nature) or piece of it is ‘theirs’ (i.e. ‘It is

mine!’) and it exists irrespective of legal ownership (Pierce et al.,

2001)

Efficacy and effectance Individuals’ judgement about their capability to perform across a variety of

situations (Bandura, 1977) Self-identity A personal cognitive connection between an individual and an object (e.g.

organisation). The individual’s perception of oneness with the target (e.g. the

organisation) (Porteous, 1976). Having a place (belonging) The extent to which an individual feels “at home” in the organisation (Porteous,

1976). Accountability The implicit or explicit expectation of the perceived right to hold others and

oneself accountable for influences on one’s target of ownership (Lerner &

Tetlock, 1999) Territoriality An individual’s behavioural expression of his/her feelings of ownership toward

a physical or social object (Brown et al. ,2005)



Concept Definition

Autonomy The regulation of the self and the extent to which a person needs or is eager to

experience individual initiative in performing in the organisation (Ryan & Deci,

2006). Responsibility The state of cognitive and emotional acceptance of responsibility (Cummings

& Anton, 1990).

4.2.2 Step 2: Generation of an item pool

For item generation the deductive process was used (Hinkin, 1998). Item generation

was thus initiated by a thorough review of the literature on possessiveness,

psychological ownership, and related terms as reported in Chapter 2. The definitions as

given in Table 4.1 were used as a guide for the development of items. Items to be

included in the measure were generated from the review of literature and expanded on

the instrument developed by Avey et al. (2009). The six theory-driven domains

determined to best constitute the dimensions of promotive or promotion-orientated

psychological ownership include self-efficacy, self-identity with the target, sense of

belonging, accountability, autonomy and responsibility. Territoriality was identified as

the seventh dimension, belonging to preventative or prevention-orientated

psychological ownership (Avey et al., 2009). The researcher generated 54 items

representing these seven theory-driven dimensions of psychological ownership. The

number of items representing each construct is indicated in Table 4.2:

Table 4.2: Items per dimension

Dimension Items Self-efficacy 7 Self-identity 8 Sense of belonging 7 Accountability 6 Territoriality 9 Autonomy 10 Responsibility 7 Total 54



A detailed theoretical verification of each item included under the seven descriptive

dimensions of psychological ownership in the questionnaire has been provided in Table

4.3.

Table 4.3: Theoretical verification of each item per dimension

SEVEN DIMENSIONS AND DESCRIPTIVE ELEMENTS OF PSYCHOLOGICAL OWNERSHIP

Element/Items Theoretical verification

A. Self-efficacy Def: Individuals’ judgment about their capability to perform across a variety of situations (Bandura, 1977)

1. I am confident that I can make suggestions about ways to improve the working of my work unit.

According to Parker (1998, p. 835), “self-efficacy concerns the extent to which people feel confident that they are able to carry out a broader and more proactive role”.

2. I have the confidence to suggest doing things differently in my work unit.

According to Bandura (as cited in Parker, 1998), perceived control is a critical determinant of self-efficacy. Andrisani (1976) argues that a high level of perceived control relates positively to personal confidence, initiative, and innate ability.

3. I am confident that I can design new procedures for my work unit/area.

Adapted from Parker’s (1998) Role Breadth Self-efficacy instrument. Original item: “How confident would you feel making suggestions to management about ways to improve the working of your section?”

4. I am confident that I am able to analyse a long-term problem to find a solution.

Adapted from Parker’s (1998) Role Breadth Self-efficacy instrument. Original item: “How confident would you feel analysing a long-term problem to find a solution?”

5. I am confident that when I make plans that will benefit the organisation, I can make them work.

Bandura (as cited in Parker, 1998) suggests that one of the four categories that are used in the development of self-efficacy is enactive mastery, or repeated performance success.

6. I am confident that I have the ability to act within the responsibilities of my job.

According to Bandura (1995), p. 193), “An efficacy expectation is the conviction that one can successfully execute the behaviour required to produce outcomes.”

7. I am confident that I can meet my performance expectations that were agreed with me upfront.

Brockner (as cited in Parker, 1998) states that self-efficacy is a judgement about specific task capability.

B. Self-identity Def: A personal cognitive connection between an individual and an object (e.g. organisation). The individual’s perception of oneness with the target (e.g. the organisation) (Porteous, 1976).

8. I personally experience the successes and failures of the organisation as my successes and failures.

“…the tendency of individuals to perceive themselves and their groups or organisations as intertwined, sharing common qualities and faults, successes and failures, and common identities” (Mael & Tetrick, 1992, p. 813). Adapted from Mael and Ashforth (1992). Original item: “This school’s successes are my successes.”





9. I feel that by identifying with the characteristics of the organisation it helps me develop a sense of who I am.

According to Pierce et al. (2003), people use ownership to define themselves, to express their self-identity to others and to maintain the continuity of the self across time.

10. I feel the need to be seen as a member of the organisation.

“…the individual defines him or herself in terms of the organisation in which he or she is a member” (Mael & Ashforth, 1992, p104).

11. It is important to me that others think highly of my organisation.

“…an individual may feel proud to be part of a group” (O’Reilly & Chatman, 1986, p. 493).

12. My personal values and that of the organisation are aligned and cared for.

“…the values of the individual and the group or organisation are the same” (O’Reilly & Chatman, 1986, p. 493).

13. It is important to me to defend my organisation to outsiders when it is criticised.

According to Lee, (1971, p. 215) cited in Edwards, 2005) where identification with the organisation is in the form of loyalty, this will relate to attitudes and behaviours that include “defending the organisation to outsiders”.

14. It is important to me to support my organisation’s goals and policies.

“The organisation’s goals become the individual’s goals, and those who identify strongly are more likely to be motivated to work hard to help achieve these goals” (Edwards, 2005, p. 207).

15. I am proud to say to every person I meet that this is my organisation.

“…an individual may feel proud to be part of a group” (O’Reilly & Chatman, 1986, p. 493) and according to Lee, as cited in Edwards, 2005, p. 215) where identification with the organisation is in the form of loyalty, this will relate to attitudes and behaviours that include “taking pride in the tenure in the organisation”.

C. Sense of belongingness Def: The extent to which an individual feels ”at home” in the workplace (Porteous, 1976).

16. I think about this organisation as MY organisation.

Adapted from the original five-item measure of psychological ownership originally developed by Pierce, Van Dyne and Cummings (1992, cited in Van Dyne & Pierce, 2004).

17. I perceive myself to be psychologically intertwined with the fate of the organisation.

“…the process by which the goals of the organisation and those of the individual become increasingly integrated or congruent” (Hall et al., 1970, p. 176).

18. I feel that I belong in this organisation. “…the perception of oneness with or belongingness to an organisation” (Mael & Ashforth, 1992, p. 104). Taken from Avey et al. (2009) Psychological Ownership Questionnaire. Original item: “I feel I belong in this organisation”.

19. I feel ”at home” in this organisation. According to Porteous (1976), it is those possessions in which an individual finds a strong sense of identification that come to be regarded as “home” – my place.

20. This organisation cares for me as a person and looks after me.

According to Pierce et al. (2003), people become psychologically attached to a variety of objects of material and immaterial nature and in many of these possessions they find a special place that is familiar and provides some form of personal security.





21. There is a strong relationship between me and my team.

“…Organisational identification is seen as a key psychological state reflecting the underlying link or bond that exists between the employee and the organisation” (Edwards, 2005, p. 201).

22. I give and receive affection from my colleagues and this bonds us with the organisation.

According to Lee (as cited in Edwards, 2005, p. 214), belongingness results from common goals shared with other employees who feel that their function fulfils their personal needs.

D. Accountability Def: The implicit or explicit expectation of the perceived right to hold others and oneself accountable for influences on one’s target of ownership (Lerner & Tetlock, 1999).

23. I will hold management accountable for their decisions.

According to Pierce et al. (2003), individuals who experience high levels of psychological ownership expect to be able to call others to account for influences on their target of ownership.

24. I have the right to hold myself and others accountable for organisational performance.

Lerner and Tetlock (1999, p. 255) refer to accountability as “the implicit or explicit expectation that one may be called on to justify one’s beliefs, feelings, and actions to others”.

25. It is important to me to have the right to information about the organisation, such as performance and projection and about my personal and team performance.

According to Pierce et al. (2001) individuals have the right to information about the target of ownership.

26. In my organisation we are allowed to make mistakes and own up to it.

“Accountability requires a level of ownership that includes: making; keeping; and proactively answering for personal commitments” (Wood & Winston, 2007, p. 168).

27. In my organisation I accept responsibility and take the consequences of these decisions.

According to Kouzes and Posner (1993, cited in Wood & Winston, 2007) accountability has to do with the acceptance of responsibility, voluntary transparency and answerability.

28. I work in an open environment where everyone is allowed to challenge a decision or strategy as long as it is done constructively.

Kubzansky and Druskat (as cited in Pierce et al., 2001) state that the right to information about the target of ownership and the right to have a voice in decisions that impact on the target are frequently associated with ownership. Adapted from Avey et al. (2009) Psychological ownership questionnaire. Item: “I would challenge the direction of my organisation to assure it’s correct”.

ETerritoriality Def: An individual’s behavioural expression of his/her feelings of ownership toward a physical or social object (Brown et al., 2005)





29. It is important to me that my organisation allows me to personalise my work space.

According to Wells, (2000) employees are generally happier if they are allowed to personalise; Brown et al. (2005, p. 581) came to the conclusion that “personalizations are an important type of marking that allow a person to express his or her identity and foster a sense of belonging to the organization”.

30. It is important to me to defend my work space from others in the organisation.

Porteous (1976) states that control over space per se is a satisfaction that is derived from ownership, and that people use control-orientated marking to persuade others not to attempt to gain access to their marked territory (Brown et al., 2005).

31. It is important to me to have a work space or work area of my own.

Porteous (1976) has argued that individuals have an inherent territoriality need, that is, a need to possess a certain space.

32. It is important to me to protect my belongings from others in the organisation.

According to Belk (1988) and Dittmar (1992), possessions can play such a dominant role in the owner’s identity that they become part of the extended self, with the result that the loss of possessions will, according to James (1890, p. 178), lead to “shrinkage of our personality, a partial conversion of ourselves to nothingness”. Item adapted from Avey et al. (2009) Psychological ownership questionnaire. Original item: “I feel I need to protect my property from being used by others in my organisation.”

33. It is important to me that people I work with do not invade my work space.

According to Brown et al. (2005), control-orientated marking communicates to others that a territory has been claimed so as to discourage access, usage, and infringement attempts by others. Item taken from Avey et al. (2009) Psychological ownership questionnaire.

34. It is important to me to protect my ideas from being used by others in the organisation.

According to Locke, (1690, cited in Pierce et al., 2001) people own their labour and, therefore, they often feel that they own that which they created, shaped or produced. Item adapted from Avey et al. (2009) Psychological ownership questionnaire. Original item: “I feel I need to protect my ideas from being used by others in my organisation.”

35. It is important to me to discourage others from attempting to enter my work space.

According to Brown et al. (2005, p. 586) “Behaviors, such as marking and defending, that increase the sense one has a place of one’s own will increase the rootedness and sense of belonging an individual member has with the organisation.”

36. It is important to me to know and have access to all policies and procedures of the organisation.

According to Pierce et al. (2001) “when employees are given information about potential targets of ownership (e.g., the mission of the organisation, its goals, and its performance), they feel that they know the organization better and, as a result, may develop psychological ownership toward it”.

37. Every person in our organisation knows the boundary of acceptable and unacceptable behaviour.

Through intimate knowledge of an object, place, or person, a union of the self with the object takes place (Beaglehole, 1932). Weil (1952) supports this by stating that people can feel that something is theirs by virtue of being associated and familiar with it.





F Autonomy Def: Refers to the regulation of the self and is the extent to which a person needs or is eager to experience individual initiative in performing a job (Ryan & Deci, 2006).

38. My job gives me the freedom to schedule my work and determine how it is done.

Autonomy reflects “the degree to which the job provides substantial freedom, independence, and discretion to the employee in scheduling the work” (Hackman and Oldham, 1975, p. 162).

39. My job allows me to have control over my working environment.

According to Ashforth and Saks (2000, p. 313), people must perceive that they have control in the working environment.

40. My job allows me to participate in making decisions that affect my task domain.

Perceived control refers to “employees’ belief about the extent to which they have autonomy in their job and are allowed to participate in making decisions on issues that effect their task domain” (Ashforth & Saks, 2000, p. 313).

41. My job allows me the opportunity for independent thought and action.

Autonomy reflects “the degree to which the job provides substantial freedom, independence, and discretion to the employee in scheduling the work” (Hackman and Oldham, 1975, p. 162).

42. My job allows me to do my work independently. Mayhew et al. (2007) suggest that organisations should provide their employees with opportunities to control facets of their employment by allowing them the freedom and flexibility to plan and perform their workloads.

43. My job allows me to use my personal initiative and judgment in carrying out my work.

Adapted from Hackman and Oldham’s (1975) Job Diagnostic survey. Original item: “The job gives me a chance to use my personal initiative and judgment in carrying out the work.”

44. My job gives me the freedom to do pretty much what I want in my job.

Deci and Ryan (1985) refer to autonomous actions as those actions that are regulated and endorsed by the self and that are accompanied by a sense of psychological freedom and violation.

45. My job gives me the freedom to act morally for the purpose of doing good for my organisation independently of incentives.

Adapted from Hackman and Oldham’s (1975) Job Diagnostic survey. Original item: “The job gives me considerable opportunity for independence and freedom in how I do the work”.

46. My job allows me to apply informed consent to my activities that I deem necessary to action my task domain.

Adapted from Hackman and Oldham’s (1975) Job Diagnostic survey. Original item: “The job gives me considerable opportunity for independence and freedom in how I do the work.”

47. My autonomy to act is restricted by the policies and procedures of the organisation but does not inhibit my ability to deliver the tasks required.

Prelinger (1959) found that the more an individual feels that he or she has control over an object, the more likely it is that that object will be perceived as part of the self.

GResponsibility Def: The state of cognitive and emotional acceptance of responsibility (Cummings & Anton, 1990).





48. I accept full responsibility for my actions within the organisation.

Mackin (as cited in Pierce et al. 2001) states that for “every right of ownership which … an owner may feel … there is a commensurate or balancing responsibility”.

49. I accept ownership for the results of my decisions and actions.

Adapted from a leader accountability instrument developed by Wood and Winston (2007). Original item: “The leader accepts responsibility for his/her actions within the organisation.”

50. I strive to contribute as much as possible to the effectiveness of the organisation.

Pierce et al. (2001) propose that a positive and causal relationship exists between the extent to which an individual employee invests himself or herself in the potential target of ownership and the degree of ownership the employee feels toward that target.

51. I feel personally responsible for the work I do in my organisation.

Dipboye (as cited in Pierce et al., 2003, p. 29) states that “When an individual’s sense of self is closely linked to the target, a desire to maintain, protect, or enhance that identity will result in an enhanced sense of responsibility.”

52. I feel I should personally take the credit or blame for the results of my work in the organisation.

Rodgers (as cited in Pierce et al., 2001, p. 303) argues that “the right to participate in decision making is balanced with an active right and responsible voice”.

53. The buck stops with me and I ensure that the task/complaint is resolved successfully every time.

Adapted from Hackman and Oldham’s (1975) four item Sense of responsibility for the job instrument. Original item: “It is hard, on this job, for me to care very much about whether or not the work got done right.”

54. If I cannot deliver on a task for whatever reason, I maintain the responsibility to find an alternative resource or solution.

Adapted from Hackman and Oldham’s (1975) four item Sense of responsibility for the job instrument. Original item: “Whether or not this job gets done right is clearly my responsibility.”

From Table 4.3 it is evident that the origins of the items included in the seven

dimensions of the questionnaire are as follows:

• Self-efficacy: Based on the self-efficacy scale developed by Parker (1998),

seven self-efficacy items were developed. Parker (1998) found an alpha

coefficient of .96 for his ten-item Role Breadth Self-Efficacy scale that was

submitted to 669 employees from a glass manufacturing company in the United

Kingdom (UK). Two of the items of Parker’s self-efficacy scale were adapted.

Central key words of the original items were retained but rephrased to fit the

structure of the newly developed questionnaire (please refer to Table 4.3). The

remaining five items were developed by the researcher, based on the theoretical

verification as displayed in Table 4.3.



• Self-identity: Eight items were developed for the measuring of self-identity. One

of these items was based on the work done by Mael and Ashforth (1992). They

measured “Organisational identification” with a six-item scale and reported

coefficient alphas from .87 to .89 in two samples of US Army squad leaders. The

remaining seven items were developed by the researcher, based on the theoretical

verification presented in Table 4.3.

• Sense of belonging: Seven sense of belonging items were compiled. One of the

items measuring sense of belonging was taken from the existing Psychological

Ownership Questionnaire developed by Avey and colleagues (2009). They found

an alpha coefficient of .92 for this particular dimension. One item, namely, “this

organisation is my organisation” was adapted from the seven-item measure of

psychological ownership developed by Van Dyne and Pierce (2004). This item was

rephrased, although the key words of the original item were retained. Cronbach’s

coefficient alpha showed acceptable internal consistency reliability respectively of

.87, .90 and .93 for three US samples. The remaining five items were developed

by the researcher, based on the theoretical verification as presented in Table 4.3.

• Accountability: Six accountability items were compiled. One item measuring

accountability was adapted from the existing Psychological Ownership

Questionnaire developed by Avey and colleagues (2009). An alpha coefficient of

.86 was reported for this particular dimension. This item was rephrased and

rewritten to form a newly developed item that would fit the structure of the newly

developed questionnaire. Based on the theoretical verification for items displayed

in Table 4.3, the researcher compiled the remaining five items.

• Territoriality: Nine items for the measuring of territoriality were compiled. Three

items measuring territoriality were adapted from the existing Psychological

Ownership Questionnaire developed by Avey and colleagues (2009). Cronbach’s

coefficient alpha showed acceptable internal consistency reliability of .83 for this

dimension. The researcher compiled the remaining six items based on the



theoretical verification for items displayed in Table 4.3

• Autonomy: Ten items for the measuring of Autonomy were compiled. Three items

from the Revised Job Diagnostic Survey of Hackman and Oldham (1975) were

adapted because they proved to be reliable items – an alpha coefficient of 0.72 for

the autonomy dimension was reported by Buys et al. (2007) on a South African

sample comprising 677 respondents from various organisations. However,

although some of the key words of the original items remained, these three items

were rephrased and rewritten to form three new items to fit the flow and structure

of the newly developed questionnaire. Seven additional items were compiled by

the researcher based on the theoretical verification provided in Table 4.3.

• Responsibility: Seven Responsibility items were compiled. Two items for the

measuring of responsibility were adapted from Hackman and Oldham’s four-item

Sense of Responsibility instrument (cited in Li, 2008). Li reported a coefficient

alpha for sense of responsibility of .79. The sample comprised 162 volunteers from

various non-profit organisations in the Waikato region of New Zealand. However,

these items were rephrased and rewritten to form two newly developed items. One

item from Wood and Winston’s (2007) Responsibility Scale was also adapted. A

sample comprising 148 employees from the US completed their questionnaire and

a remarkably high coefficient alpha score of .97 was reported. Although some of

the key words of the original item remained, the item was rephrased to fit the

current structure of the newly developed questionnaire. Four additional items were

developed by the researcher (see the theoretical verification provided in Table

4.3).



4.2.3 Step 3: Determining the format of the questionnaire

A Likert-type rating scale with an equal 1-6 agreement format was chosen, where:

1 = Strongly disagree

2 = Disagree

3 = Slightly disagree

4 = Slightly agree

5 = Agree

6 = Strongly agree

The Likert-type scale was chosen above a typical dichotomous “yes-no or true-false”

scale because Likert-type scales are most frequently used in survey questionnaire

research and are most useful in behavioural research, according to Hinkin (1998). The

Likert-type scale allows respondents to indicate their degree of agreement with the

particular statement (DeVellis, 2003). A desirable quality of a measurement scale is to

generate sufficient variance among respondents for subsequent statistical analyses. It

was noted that an equal number of options could result in respondents’ falling to one

side; however, the mid-range option of three in the scale could lead to respondents

choosing the middle option.

4.2.4 Step 4: Having the initial pool reviewed by a panel of experts and pilot

study

According to De Vos (2002), content validity is concerned with the sampling adequacy

or representativeness of the content of an instrument, thus: “Does the instrument

address whether items on an instrument adequately measure a desired domain of

content?” (Grant & Davis, 1997) In order to determine the content validity of the

psychological ownership measure, which assisted in the retention or rejection of certain

items, Lawshe’s (1975) content validity technique was applied. This judgement-

quantification process entails asking a specific number of subject matter experts to



evaluate the validity of items individually, as well as the entire instrument. The experts

had to meet a predetermined set of five criteria in order for them to be regarded as

Subject Matter Experts for the purpose of this study. He or she must

• have at least a three-year degree in the fields of industrial psychology or psychology,

human resource management or related field

• have at least five years’ work experience and expertise in applied psychology or

related fields

• have had at least one article published in a refereed journal or have presented a

paper at an international conference

• be registered with a professional body such as the Health Professions Council of

South Africa (HPCSA), South African Board of People Practice (SABPP) or

equivalent

• be regarded as an expert in the field of applied psychology or related fields by his or

her colleagues and clients.

Questionnaires were distributed to the group of experts and they were requested to

indicate whether or not a measurement item in a set of other measurement items was

essential to the functionality of the construct. They were also asked to provide

biographical information such as: their highest qualification, work experience in applied

psychology or related fields, whether or not they were registered with a professional

board, the number of their publications in refereed journals and papers presented at

international conferences and their age. This information was used to determine

whether or not each respondent did fill the criteria for a subject matter expert that had

been set for the purposes of this study. Table 4.4 indicates how the respondents met

such qualifying criteria.



Table 4.4: Subject matter expert criteria

Respondent No

Minimum 3-year degree in HR/ Industrial Psych/ Psych

or related

Minimum 5 years’ work

experience

Registered with

professional body or

equivalent

Regarded as expert by

clients/ colleagues

Publications in refereed journals /

Age Meet SME criteria?

1 DPhil

(Industrial

Psychology)

38 HPCSA

SABPP

Yes 31 64 Yes

2 PhD (HRM) 30 SANRF B-

rated

researcher

Yes 30 62 Yes

3 DPhil 28 HPCSA Yes 51 Yes

4 PhD

(Organisational

Behaviour)

25 SABPP Yes 6 53 Yes

5 PhD

(Organisational

Behaviour)

25 Chairperson:

Centre of I/O

Psychology

advisory

committee

(UNISA)

Yes 8 49 Yes

6 MA (Research

Psychology)

8 HPCSA Yes 0 32 Yes

7 D Com (HRM) 15 HPCSA

SABPP

Yes 13 50 Yes

8 MPhil (HRM) 7 Academy of

Management

Yes 1 33 Yes

9 MA

(Counselling

Psychology)

35 HPCSA Yes 10 62 Yes

It is clear from the information in Table 4.4 that all of the subject matter experts are

indeed experts in evaluating the construct, as nine had obtained a minimum of a

Master’s degree in the field of industrial psychology, human resource management or

related field. Six have doctoral degrees, whereas two of the three with Master’s degrees

are currently enrolled as doctoral students in organisational behaviour. Six of the

experts have at least 25 years of work experience and the remaining three a minimum



of seven years. They are all registered with a professional body such as the Health

Professions Council of South Africa (HPCSA) as psychologists or industrial

psychologists or at the South African Board of People Practice (SABPP) as Master

Human Resource Practitioners. One of the respondents is registered as a B-rated

researcher at the South African National Research Foundation (SANRF). The remaining

two serve as either a chairperson or member on various other human resource related

committees. They are all regarded as experts in the field of applied psychology by their

clients and colleagues. With an average age of 51, the assumption can be made that

the experts are a very experienced group of people. With the exception of one, all these

experts had had numerous articles published in refereed journals. Although one of the

respondents did not have any articles published in a refereed journal, eight papers had

been presented at international conferences. All of these experts are experienced in the

field of scale development and quite a number of them have developed a measuring

instrument as part of their study for their doctoral thesis. Thus, all nine of the

respondents met the predetermined criteria that had been set and qualified as a subject

matter expert.

The panel was asked not only to judge each item related to the specific dimension of

psychological ownership, but to indicate the clarity of each item as well as to comment

on the comprehensiveness of the entire instrument and addition of items.

Item content: the experts were provided with the conceptual definition of

psychological ownership and the relevant dimensions of psychological ownership: self-

efficacy, self-identity, sense of belongingness, accountability, territoriality, autonomy

and responsibility (Addendum A).

Lawshe’s (1975) quasi-quantitative approach to content validity was also used to

facilitate the retention or rejection of specific items. The Content Validity Ratio (CVR) for

each item was computed by making use of the following formula:


CHAPTER 4 – RESULTS AND FINDINGS 8

Where:

ne = number of subject matter experts who indicated the item as essential;

N = the total number of subject matter experts on the panel

The CVR formula takes on values between -0.1 (where none of the experts think that

the particular item is essential) and +0.1 (where all the experts regard that particular

item as essential). A CVR-value of 0 (CVR=0), indicates that 50% of the experts in the

selected panel of size N (N = number of experts) believe that the measurement item is

essential. A CVR of > 0.0 will thus indicate that more than half of the subject matter

experts believe that the particular item is essential.

The CVR is negative if fewer than half of the experts indicate that an item is essential

and positive when more than half of the experts indicate it is essential. Hence the more

experts over 50% that perceive the item as essential, the greater the degree of its

content validity. A guideline to use as minimum CVR for different panel sizes based on

a one-tailed test at the � = .05 significance level was established by Lawshe (1975).

A total of nine subject matter experts completed the questionnaire. The minimum CVR

values, according to the panel size, for an item to be retained as part of the content

validity testing (Lawshe, 1975) was .78. All items with a CVR value of less than .78

should be rejected. Lawshe’s content validity results are presented in Table 4.5.



Table 4.5: Lawshe’s content validity results

DIMENSIONS and descriptive elements of psychological ownership

Endorsement of statements CVR Retain (yes / no)

Element Not essential

Essential Item is

clear

Item is unclear

A. Self-efficacy Def: The individual’s judgment about their capability to perform across a variety of situations.

1. I am confident that I can make suggestions about ways to improve the working of my work unit.

1 8 8 1 .78 Yes

2. I have the confidence to suggest doing things differently in my work unit.

1 8 9 0 .78 Yes

3. I am confident that I can design new procedures for my work unit/area.

1 8 8 1 .78 Yes

4. I am confident that I am able to analyse a long-term problem to find a solution.

3 6 5 4 .33 No

5. I am confident that when I make plans that will benefit the organisation, I can make them work.

1 8 9 0 .78 Yes

6. I am confident that I have the ability to act within the responsibilities of my job.

2 7 9 0 .56 No

7. I am confident that I can meet my performance expectations that were agreed with me upfront.

0 9 9 0 1 Yes

B. Self-identity Def: A personal cognitive connection between an individual and an object (e.g. organisation). The individual’s perception of oneness with the target (e.g. the organisation).

8. I personally experience the successes and failures of the organisation as my successes and failures.

0 9 9 0 1 Yes

9. I feel that by identifying with the characteristics of the organisation it helps me develop a sense of who I am.

1 8 6 3 .78 Yes

10. I feel the need to be seen as a member of the organisation.

0 9 8 1 1 Yes

11. It is important to me that others think highly of my organisation.

0 9 9 0 1 Yes

12. My personal values and that of the organisation are aligned and cared for.

2 7 5 4 .56 No

13. It is important to me to defend my organisation to outsiders when it is criticised.

0 9 8 1 1 Yes

14. It is important to me to support my organisation’s goals and policies.

0 9 8 1 1 Yes

15. I am proud to say to every person I meet that this is my organisation.

0 9 8 1 1 Yes

C. Sense of belongingness Def: The extent to which an individual feels ‘at home’ in the work place.






Essential Item is

clear

Item is unclear

16. I think about this organisation as MY organisation.

0 9 8 1 1 Yes

17. I perceive myself to be psychologically intertwined with the fate of the organisation.

1 8 5 4 .78 Yes

18. I feel that I belong in this organisation. 0 9 9 0 1 Yes 19. I feel ‘at home’ in this organisation. 1 8 9 0 .78 Yes 20. This organisation cares for me as a

person and looks after me. 2 7 9 0 .56 No

21. There is a strong relationship between me and my team.

3 6 6 3 .33 No

22. I give and receive affection from my colleagues and this bonds us with the organisation.

4 5 4 4 .111 No

D. Accountability Def: The implicit or explicit expectation of the perceived right to hold others and oneself accountable for influences on one’s target of ownership.

23. I will hold management accountable for their decisions.

2 7 4 5 .56 No

24. I have the right to hold myself and others accountable for organisational performance.

0 9 4 5 1 Yes

25. It is important to me to have the right to information about the organisation, such as performance and projection and about my personal and team performance.

1 8 4 5 .78 Yes

26. In my organisation we are allowed to make mistakes and own up to it.

2 7 8 1 .56 No

27. In my organisation I accept responsibility and take the consequences of these decisions.

0 9 7 2 1 Yes

28. I work in an open environment where everyone is allowed to challenge a decision or strategy as long as it is done constructively.

3 6 8 1 .33 No

E. Territoriality Def: An individual’s behavioural expression of his/her feelings of ownership toward a physical or social object.

29. It is important to me that my organisation allows me to personalise my work space.

2 7 7 2 .56 No

30. It is important to me to defend my work space from others in the organisation.

2 7 7 2 .56 No

31. It is important to me to have a work space or work area of my own.

1 8 7 2 .78 Yes






Essential Item is

clear

Item is unclear

32. It is important to me to protect my belongings from others in the organisation.

4 5 7 2 .11 No

33. It is important to me that people I work with do not invade my work space.

2 7 7 2 .56 No

34. It is important to me to protect my ideas from being used by others in the organisation.

2 7 6 3 .56 No

35. It is important to me to discourage others from attempting to enter my work space.

4 5 5 2 .11 No

36. It is important to me to know and have access to all policies and procedures of the organisation.

3 6 8 1 .33 No

37. Every person in our organisation knows the boundary of acceptable and unacceptable behaviour.

2 7 8 1 .56 No

F. Autonomy Def: Refers to the regulation of the self and is the extent to which a person needs or is eager to experience individual initiative in performing a job.

38. My job gives me the freedom to schedule my work and determine how it is done.

0 9 7 2 1 Yes

39. My job allows me to have control over my working environment.

1 8 6 3 .78 Yes

40. My job allows me to participate in making decisions that affect my task domain.

1 8 6 3 .78 Yes

41. My job allows me the opportunity for independent thought and action.

0 9 7 2 1 Yes

42. My job allows me to do my work independently.

1 8 7 2 .78 Yes

43. My job allows me to use my personal initiative and judgment in carrying out my work.

0 9 7 2 1 Yes

44. My job gives me the freedom to do pretty much what I want in my job.

1 8 6 3 .78 Yes

45. My job gives me the freedom to act morally for the purpose of doing good for my organisation independently of incentives.

3 6 6 3 .33 No

46. My job allows me to apply informed consent to my activities that I deem necessary to action my task domain.

5 4 2 4 -.11 No

47. My autonomy to act is restricted by the policies and procedures of the organisation but does not inhibit my ability to deliver the tasks required.

1 8 5 4 .78 Yes

G. Responsibility Def: The state of cognitive and emotional acceptance of responsibility.






Essential Item is

clear

Item is unclear

48. I accept full responsibility for my actions within the organisation.

0 9 9 0 1 Yes

49. I accept ownership for the results of my decisions and actions.

0 9 9 0 1 Yes

50. I strive to contribute as much as possible to the effectiveness of the organisation.

3 6 6 3 .33 No

51. I feel personally responsible for the work I do in my organisation.

1 8 9 0 .78 Yes

52. I feel I should personally take the credit or blame for the results of my work in the organisation.

0 9 9 0 1 Yes

53. The buck stops with me and I ensure that the task / complaint is resolved successfully every time.

1 8 8 1 .78 Yes

54. If I cannot deliver on a task for whatever reason, I maintain the responsibility to find an alternative resource or solution.

0 9 9 0 1 Yes

Average number of endorsements 1.315 7.593 .72

According to the results, the majority of measurement items with regard to each

dimension were valid, since their CVR values were greater than or equal to .78 at a

significance level of � = .05, except for the dimensions of Sense of belonging (where 3

out of the original 6 items had to be rejected) and Territoriality (where 8 out of the

original 9 items had to be rejected).

Rejection of such a large number of the Territoriality items could be due to the fact that

the experts experience territoriality as negative and as a potential threat to

psychological ownership. The researcher purposefully did not disclose to the experts

that territoriality is a preventative form of psychological ownership. Their response to

these items is a confirmation of the researcher’s view that territoriality is a preventative

form of psychological ownership. In their study, Brown et al. (2005. p. 580) focused on

the territoriality concept as being behavioural and proposed that “the stronger an

individual’s psychological ownership of an object, the greater the likelihood he or she

will engage in territorial behaviours”. Although Pierce et al. (2001) argue that

psychological ownership is a cognitive-affective construct, Avey et al. (2009) focus

heavily on the cognitive aspects (versus behavioural displays) of territoriality as a



preventative form of psychological ownership. This also applies to this study. Although

territoriality may lead people to become too preoccupied with their “objects of

ownership” at the expense of their performance or other pro-social behaviours (Avey et

al., 2009), the possibility exists that feelings of territoriality may promote positive

organisational outcomes. If individuals believe that by protecting their territory they are

doing what is right (Altman, as cited in Avey et al., 2009), territoriality may lead to

increased retention and performance. Scholars such as Porteous (1976) have

suggested that individuals exercise control by the “marking” of objects, which

contributes to their attachment to the object and experienced psychological ownership.

This type of behaviour may cause the individual to feel more secure and “at home” and

they may feel that they discover themselves in the marked object. This study supports

the viewpoint of Avey et al. (2009) that territorial psychological ownership with its

typically negative implication may have a positive side.

A total of 20 items were rejected because the CVR values were less than .78. Although

item 4 (part of self-efficacy dimension), item 12 (part of self-identity dimension) and item

23 (part of the accountability dimension) had CVR values of less than .78, it was

decided to retain these three items since quite a few experts indicated that these items

were unclear and that if the questions could be rephrased, they could be retained. The

valid items from each dimension were retained. The number of items retained after the

application of Lawshe’s (1975) technique was 34. Table 4.7 gives a summary of the

original number of items compared with the number of items retained after the

application of Lawshe’s (1975) technique.



Table 4.6: Comparison between the original number of items and items retained after the application of Lawshe’s technique

Dimension Original number of items

Items retained

Self-efficacy 7 5 Self-identity 8 7 Belonging 7 4 Accountability 6 3 Territoriality 9 1 Autonomy 10 8 Responsibility 7 6 Total 54 34

Item style: The subject matter experts were also asked to evaluate each item’s

clarity and conciseness. Unclear (vague) items were indicated, as can be seen in Table

4.6. In some cases the experts suggested alternative wordings of the questions. Items

were reworded and clarified accordingly, as can be seen in Table 4.8.

Comprehensiveness: The panel of experts were also asked to evaluate the

total instrument for comprehensiveness. The panel all agreed that all the dimensions of

the desired content domain of the psychological ownership concept had been included.

However, in judging the entire instrument, the panel of experts suggested that additional

items should be added in order to represent the total content domain. The inclusion of

additional items in the questionnaire would also help in determining the validity of the

final scale, as suggested by Worthington and Whittaker (2006). Idaszak et al. (1988)

support this by stating that an instrument should have at least four to six items per scale

because this will increase the likelihood that a factor analysis will accurately reflect the

true underlying structure of the item pool. Therefore, in the second round of items

derived from the literature study 24 additional items as per Table 4.7 were added to

each one of the dimensions to better represent the total content domain.



Table 4.7: Additional items as per seven dimensions


Additional Elements / Items Theoretical verification

A. Self-efficacy Def: The individuals’ judgement about their capability to perform across a variety of situations (Bandura, 1977)

1. I am confident in my ability to execute the required tasks of my job.

According to Bandura (1977, p. 193), “An efficacy expectation is the conviction that one can successfully execute the behavior required to produce outcomes”.

2. I am confident that I can implement policies and procedures in my work environment.

According to Bandura (1977, p. 193), “An efficacy expectation is the conviction that one can successfully execute the behavior required to produce outcomes”.

3. I feel that I can represent my work environment with all internal / external stakeholders.

Adapted from Parker’s (1998) Role Breadth Self-efficacy instrument.

4. I am confident to act as an expert in my field for my work environment.

According to Bandura (1986, cited in Parker, 1989), perceived control is a critical determinant of self-efficacy. Andrisani (1976) argues that a high level of perceived control relates positively to personal confidence, initiative, and innate ability.

B. Self-identity Def: A personal cognitive connection between an individual and an object (e.g. organisation). The individual’s perception of oneness with the target (e.g. the organisation) (Porteous, 1976).

5. I act to the benefit of my organisation. Having a membership that shares the organisation’s goals and values can ensure that individuals act instinctively to benefit the organisation” (O’Reilly & Chatman, 1986, p. 493).

6. I feel part of the larger organisational entity. “…individuals perceive themselves to be part of a larger organisation” (Rousseau, 1998, p. 217).

7. I feel a strong linkage between me and my organisation.

“…the individual … see him or herself as psychologically intertwined with the fate of the group” (Mael & Ashforth, 1992, p 105).

C. Sense of belongingness Def: The extent to which an individual feels “at home” in the work place (Porteous, 1976).

8. I feel totally comfortable being in the organisation. According to Brown, (1969, cited in Edwards, 2005) the basic components of organisational identification are: attraction to the organisation, consistency of organisational and individual goals, loyalty, and reference to the self to organisational membership.



9. I feel that this organisation is part of me “…the perception of oneness with or belongingness to an organisation” (Mael & Ashforth, 1992, p. 104) and according to Rousseau (1998) the organisation becomes a part of the individual’s self-concept.

10. I feel I have a considerable emotional investment in my organisation.

“…the process by which the goals of the organisation and those of the individual become increasingly integrated or congruent” (Hall et al., 1970, p. 176).

11. I feel I have a strong bond with the organisation. “…Organisational identification is seen as a key psychological state reflecting the underlying link or bond that exists between the employee and the organisation” (Edwards, 2005, p. 201)

12. I feel secure in this organisation. Dittmar (1992) believes that possessions may provide a sense of security, and according to Porteous (1976), ”the home” is important because it provides the individual with both psychic and physical security.

D. Accountability Def: The implicit or explicit expectation of the perceived right to hold others and oneself accountable for influences on one’s target of ownership (Lerner & Tetlock, 1999).

13. I would take action against inappropriate behaviour in my organisation.

In their stewardship theory, Davis et al. (1997) propose that in certain situations when individuals feel like stewards, they will be motivated to act in the best interest of the principals rather than in their personal interests. Pierce et al. (2003, p. 30) thus came to the conclusion that “when individuals feel psychological ownership, they may feel as though they are the ‘psychological principals’ or stewards and act accordingly”.

14. I would challenge a decision or strategy being made in the organisation.

Kubzansky and Druskat (1993, cited in Pierce et al., 2001) state that the right to information about the target of ownership and the right to have a voice in decisions that impact the target are frequently associated with ownership. Adapted from Avey et al. (2009) Psychological ownership questionnaire. Item: “I would challenge the direction of my organisation to ensure it’s correct.”

15. I would report inappropriate behaviour in my organisation.

Adapted from Avey et al. (2009) Psychological ownership questionnaire. Item: “I would not hesitate to tell my organisation if I thought something was done wrong.”

16. I acknowledge my mistakes in the organisation. “Accountability requires a level of ownership that includes: making; keeping; and proactively answering for personal commitments” (Wood & Winston, 2007, p. 168).

17. I take responsibility for my decisions in the organisation.

According to Kouzes and Posner (1993, cited in Wood & Winston, 2007) accountability has to do with the acceptance of responsibility, voluntary transparency and answerability.



18. I hold myself and others accountable for organisational performance. This question was split into two, due to the fact that it was double-barrelled. I hold myself… and I hold others…

ETerritoriality Def: An individual’s behavioural expression of his/her feelings of ownership toward a physical or social object (Brown, et al., 2005)

No additional items were added. Original items remained, as was discussed in paragraph 4.2.4

FAutonomy Def: Refers to the regulation of the self and is the extent to which a person needs or is eager to experience individual initiative in performing a job (Ryan & Deci, 2006)

19. I have almost complete responsibility for deciding how and when the work is done.

According to Pierce O’Driscoll et al. (2004), the creating and maintaining of work settings that empower individuals and enable them to exercise control over important aspects of their work arrangements should enhance their sense of ownership, which may promote the manifestation of work-related attitudes and behaviours.

20. I have considerable opportunity for independence and freedom in how I do my work.

Adapted from Hackman and Oldham’s (1975) Job Diagnostic survey. Original item: “The job gives me considerable opportunity for independence and freedom in how I do the work.”

GResponsibility Def: The state of cognitive and emotional acceptance of responsibility (Cummings & Anton, 1990).

21. I would invest time and energy beyond my job in the organisation

Pierce et al. (2001) propose that a positive and causal relationship exists between the extent to which an individual employee invests himself or herself into the potential target of ownership and the degree of ownership the employee feels toward that target.

22. I proactively enhance both tangible and intangible targets of my organisation

The feelings of ownership toward both material and immaterial objects can not only shape identity (as was mentioned earlier by Belk (1988) and Dittmar (1992)), but can also affect behaviour (Isaacs, 1933).

23. I would protect, care and nurture all elements of my organisation

Pierce et al. (2003, p. 29) state that “Psychological ownership for a particular target may also promote feelings of responsibility that include feelings of being protective, caring, and nurturing and the proactive assumption of responsibility for that target.”

Five of the originally nine experts served as a second set of expert judges for content

validation of the remaining 69 items and agreed that all 69 items be included in the final

construct measure of psychological ownership. Table 4.8 indicates the number of items



to be included in the final instrument after the application of Lawshe’s technique and

after additional items had been added.

Table 4.8: Number of items included in the final instrument after additional items had been added

Dimension Number of items retained after

the application of Lawshe’s technique

Number of items kept despite

Lawshe’s application

Second round of items derived from

literature study

Total number of items to be

included in the final

instrument to be tested

Self-efficacy 5 1 4 10 Self-identity 7 1 4 11 Belonging 4 5 9 Accountability 3 1 6 10 Territoriality 1 8 9 Autonomy 8 2 10 Responsibility 6 3 9 Total 34 11 24 69

Figure 4.2 summarises in a flow diagram the development process in the final construct

measure as reflected in Table 4.8.

Figure 4.2: Development process of items

$��!

��

(��!

��

��-

��! �

��.��

�((��

=�- �+�/�

��+��2 �

>1� ��?@1

"��

��

! ��

1>A�1?BC

��! ��(�

��(��#��

��.��#

=�- �+�/�

�((��

@1A��?1>

D ! #��

��- ��! �

��

��!

��

��-

08

6?

28

8778



• Pilot study

A pilot study was initiated. The purpose of the pilot study was twofold: firstly, to test the

experimental process to be employed in the study and, secondly, to get an indication of

how the measures kept together. A questionnaire consisting of the 69 psychological

ownership items was administered via paper and pencil to a small group of individuals

(N = 46) from the same population as that for which the eventual project was intended.

The pilot study ended with the participants’ completion of a combination of measures

which included the psychological ownership, affective commitment, turnover intentions

and job satisfaction measures. The affective commitment, turnover intentions and job

satisfaction measures were included to examine the nomological network of the variable

of interest as part of the construct validation process (Judge, Erez, Bono & Thoresen,

2003).

Although the pilot sample was small, it was felt that conducting the pilot study would

help to identify ambiguous or unclear items, as well as ease of completion of the

questionnaire (Welman & Kruger, 1999). The outcome of the pilot study could also offer

some insight into the potential reliability of not only the psychological ownership

measure, but also the other measures (Hess, as cited in Faranda, 2001). Subsequent to

the completion of the questionnaire, respondents were requested to comment on the

clarity of the items.

• Preliminary analysis on pilot study

Each of the dimensions of the psychological ownership measure achieved a satisfactory

reliability coefficient. Several respondents, however, expressed confusion regarding the

meaning of the words “own up to” used in items 8 and 16. As a result, the wording was

changed as indicated in Table 4.9.



Table 4.9: Original versus revised items identified from the pilot study

Item no Original item Revised item

8 I own up to my mistakes in the organisation.

I acknowledge my mistakes in the organisation

16 I own up to the consequences of my decisions in the organisation.

I accept the consequences of my decisions in the organisation

The alpha values for the respective dimensions of the psychological ownership measure

for the pilot study are indicated in Table 4.10.

Table 4.10: Alpha values for pilot study per dimension

Seven Dimensions Alpha coefficient

Self-efficacy .83

Self-identity .88

Sense of belonging .90

Accountability .72

Territoriality .60

Autonomy .92

Responsibility .87

With respect to exploratory measurement research, these alpha values surpass the

moderate reliabilities of .50 – .60 that were suggested by Nunnaly (1967). For basic

research, according to Peter (1979), values close to .80 are definitely adequate, while

Carmines and Zeller (1979) prefer alphas above .80. Pilot-study alpha values for the

other measures taken were, according to the criteria mentioned, also adequate. The

coefficient alpha for affective commitment was .69. For turnover intentions and job

satisfaction, alphas were .89 and .60, respectively.

The final psychological ownership instrument comprising 69 items can be viewed in

Annexure B.



4.2.5 Step 5: Administering items to a development sample

A non-probability convenience sample of 712 was collected from employed

professional, high-skilled and skilled individuals in various organisations in both the

private and public sector in South Africa. If sample size permits, the sample may be

randomly split into two subsets (Hair et al., 2006). The reason for this split in this study

was due to the fact that data were collected at one time. One half of the sample was

used for the development of a model and the other half of the sample was used to

validate the results that were obtained from the first half (Anderson & Gerbing, 1988).

Thus, a sample of 356 respondents was used for Exploratory Factor Analysis (EFA) and

a sample of 356 for Confirmatory Factor Analysis (CFA). The sample size is in

accordance with the guideline established by Worthington and Whittaker (2006), that

300 is generally sufficient for factor analysis. The ratio of 5.1:1 of the sample size to the

number of variables met the guideline set by Hair et al. (2006), who suggest that the

number of observations per variable should be a minimum of five and hopefully at least

ten observations per variable.

4.2.6 Step 6: Initial item reduction

4.2.6.1 Exploratory factor analysis (EFA)

According to Worthington and Whittaker (2006, p. 807), the main purpose of the EFA is

to “group a large item set into meaningful subsets that measure different factors”. An

EFA was conducted to determine the following: (1) the number of factors that underlie

the set of items and (2) to define the underlying dimensionality of the set of items

(Tabachnick & Fidell, 2007). This would enable the researcher to identify those items

that did not measure an anticipated factor or that simultaneously measured multiple

factors. These items could be poor indicators of the preferred construct and could be

eliminated from further research.



Before the commencement of an EFA, it is important to determine the factorability of the

correlation matrix. As indicated in Table 4.11, a statistically significant Bartlett‘s Test of

Sphericity (sig. < .50) showed that sufficient correlations existed among the variables to

proceed with a factor analysis. The KMO measure of sample adequacy of 0.931, which

is well above the guideline of .60 (Tabachnick & Fidell, 2007) confirmed that the overall

significance of the correlations within the correlation matrix was suitable for factor

analysis.

Table 4.11: KMO and Bartlett's Test results

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .931

Bartlett's Test of

Sphericity

Approx. Chi-Square 10445.178

Df 1326

Sig. .000

In the EFA the responses on the 69 items of the Psychological Ownership

Questionnaire (POSQ) were correlated and rotated using maximum-likelihood factor

extraction with oblique rotation (direct oblimin, delta = 0) using SPSS statistical

software. Maximum-likelihood factoring estimates the factor loadings for the population

that maximise the likelihood of sampling the observed correlation matrix (Tabachnick &

Fidell, 2007). In this study an oblique rotation was employed because the factors in the

psychological ownership measure were considered to be correlated.

The scree plot and parallel analysis was used to determine the number of factors to be

considered as suitable for further retention. The scree plot and parallel analysis in

Figure 4.3 indicated that only four significant factors from the originally defined seven

factors could be identified from the 69 items.



Figure 4.3: Scree plot of the actual and the random data for 69 items

In the parallel analysis a break can be observed on the scree plot between factors four

and five. The eigenvalues of the random data set (the solid line) intersect the

eigenvalues for the actual data (dotted line) set at factor five, signifying four significant

factors. The results reported in Table 4.12 indicate that four significant factors explain

only 44.79% of the total variance. According to Hayton et al. (2004), as many common

factors as possible should be kept to explain at least 50% of the variance in the data

set.

Table 4.12: Factor eigenvalues and variance explained for the 69 items

Factor Total % of variance Cumulative %

1 19.652 28.481 28.481

2 4.976 7.212 35.693

3 3.432 4.974 40.667

4 2.847 4.126 44.794

5 2.003 2.903 47.697

6 1.796 2.603 50.300

7 1.512 2.192 52.491

8 1.296 1.878 54.369

9 1.286 1.863 56.232

10 1.243 1.801 58.034




11 1.152 1.670 59.703

12 1.120 1.623 61.326

13 1.088 1.577 62.903

14 1.011 1.466 64.369

15 .973 1.410 65.778

16 .967 1.401 67.180

17 .914 1.324 68.504

18 .895 1.297 69.801

19 .845 1.225 71.026

20 .815 1.181 72.207

21 .797 1.155 73.362

22 .784 1.136 74.498

23 .707 1.025 75.523

24 .684 .992 76.515

25 .681 .987 77.501

26 .666 .966 78.467

27 .637 .924 79.391

28 .626 .908 80.298

29 .609 .883 81.181

30 .574 .832 82.013

31 .549 .796 82.809

32 .520 .754 83.563

33 .512 .742 84.305

34 .503 .729 85.034

35 .482 .698 85.732

36 .471 .683 86.415

37 .457 .663 87.078

38 .449 .650 87.729

39 .436 .631 88.360

40 .433 .627 88.987

41 .419 .607 89.594

42 .398 .577 90.171

43 .385 .558 90.729

44 .376 .545 91.274

45 .367 .532 91.806

46 .353 .512 92.318




47 .333 .483 92.801

48 .328 .475 93.276

49 .322 .467 93.743

50 .314 .455 94.198

51 .304 .440 94.638

52 .291 .421 95.059

53 .273 .395 95.455

54 .263 .382 95.836

55 .252 .365 96.202

56 .241 .350 96.551

57 .237 .344 96.895

58 .226 .328 97.223

59 .220 .318 97.542

60 .214 .310 97.852

61 .204 .295 98.147

62 .194 .281 98.428

63 .189 .274 98.702

64 .177 .257 98.959

65 .159 .230 99.189

66 .148 .215 99.403

67 .144 .209 99.612

68 .139 .202 99.814

69 .128 .186 100.000 Extraction Method: Maximum Likelihood.

Kaiser’s criterion, compared with Cattell’s scree test and parallel analysis, clearly

overestimated the number of true factors for the data set.

In the first round of Exploratory Factor Analysis on the four-factor model, all items with

factor loadings of less than .32 in the rotation matrix were removed (Worthington &

Whittaker, 2006). Items that cross-loaded were deleted as well. Only 52 items were

retained and they were subjected to a second round of Exploratory Factor Analysis.



In the second round of Exploratory Factor Analysis on the four-factor model an

additional two items that loaded significantly across two or more factors were omitted.

To create a more parsimonious and simple structure, only those items that clearly

loaded on a single appropriate factor, and were based on the researcher’s

understanding of the theoretical foundation of the construct measured, were retained.

Therefore an additional 12 items were omitted that were theoretically inconsistent with

their factor. For example, two responsibility items, four accountability items, one

territoriality item and one self-efficacy item that loaded on the belongingness factor were

omitted. Two self-efficacy and one self-identity item that loaded on the responsibility

factor and one self-efficacy and one territoriality item that loaded on the autonomy factor

were deleted as well.

Self-efficacy items

Although ten items were written to capture the dimension of self-efficacy, none of these

items survived the stages of scale development. The items either cross-loaded or

loaded on dimensions which were theoretically inconsistent with the factor. Control of

objects leads to perceptions of personal efficacy. According to Furby (1978),

possessions came to be part of the extended self and are therefore important to the

individual because they are instrumental in exercising control over the physical

environment as well as over people. Control is a key characteristic of the phenomenon

of ownership. The greater the amount of control, the more the object is experienced as

part of the self.

The individual’s self-concept is strongly influenced by culture. In their study Janse van

Rensburg and Roodt (2005) found that race groups differ in their perceptions of

employment equity (EE) and black economic empowerment BEE). These perceptions

have strong bearing on people’s beliefs, values and needs. In another South African

study, Urban (2006) found that White South Africans had lower mean scores with

regard to self-efficacy comparing to Indian and Black South Africans.



It might be that the White respondents that account for 60% of the sample do have the

perception that due to EE and BEE they are losing control over their environment as

well as over people.

Self-identity and sense of belonging items

In the four-factor solution, ten of the self-identity items and eight of the sense-of-

belonging items loaded on one factor. The researcher decided to retain these items as

part as one dimension because these constructs are essentially the same; according to

Lee (as cited in Edwards, 2005, p. 210), identification involves a sense of

belongingness that results “from common goals shared with others in the organisation

or as a result of employees’ feeling that their function within the organisation is

important in fulfilling their personal needs”. Ashforth and Mael (1989, p. 21) refer to

identification as “the perception of oneness or belongingness to some human

aggregate”, or “when a person’s self-concept contains the same attributes as those in

the perceived organizational identity” (Dutton, Dukerich & Harquail, 1994).

Accountability items

Ownership is frequently associated with a bundle of rights (Pierce et al, 2001). Most

frequently associated with ownership are the right to information about the target of

ownership and the right to have a voice in decisions that impact the target.

The expectation of information sharing and permission to influence the direction of the

target are consequences of the right to hold others accountable. It might be that the

White respondents in the sample that account for 60% of the sample do have the

perception that they’ve lost their right to have a voice in the workplace due to the

implementation of EE and BEE.



Responsibility items

The remaining two of the original ten accountability items loaded on the responsibility

factor. According to Bavly (as cited in Wood & Winston, 2007), accountability implies the

acceptance of responsibility. It seems that the sample respondents interpreted these

two questions as part of their responsibility, accepting the responsibility for them, rather

than as belonging to the accountability factor.

Territoriality items

Only five of the original nine territoriality items survived the stages of scale

development. The remaining items either cross-loaded on other factors or loaded on

dimensions which were theoretically inconsistent with the factor.

Therefore the researcher decided to retain two accountability items but as part of the

responsibility factor. Only 35 of the original 69 items now remained. Once more

Bartlett’s Test of Sphericity [χ2 (703) =7783.467, p<.001] and the KMO measure of sample

adequacy (0.923) pointed out that the attributes of the correlation matrices of the 35

item scores would probably factor well.

The scree plot and parallel analysis shown in Figure 4.4 indicates once again that only

four significant factors could be identified. In the parallel analysis a clear break can be

observed on the scree plot between factors four and five. The eigenvalues of the

random data set (the solid line) intersect the eigenvalues for the actual data (dotted line)

set at factor five, signifying four significant factors.



Figure 4.4: Scree plot of the actual and the random data for 35 items

The results reported in Table 4.13 indicated that the four factors of the SAPOS,

comprising 35 items, explained 57.37% of the total variance. This is in accordance with

the recommendation of Hayton et al. (2004) that as many common factors as possible

should be kept to explain at least 50% of the variance in the data set.

Table 4.13: Factor eigenvalues and variance explained for the 35 items


1 11.708 33.452 33.452

2 3.445 9.842 43.294

3 2.622 7.492 50.786

4 2.305 6.586 57.372

5 1.122 3.205 60.576

6 1.010 2.886 63.463

7 .944 2.697 66.160

8 .884 2.526 68.685

9 .798 2.280 70.965

10 .719 2.054 73.020

11 .687 1.962 74.982




12 .627 1.791 76.773

13 .600 1.715 78.488

14 .563 1.609 80.097

15 .546 1.561 81.658

16 .495 1.415 83.073

17 .471 1.346 84.419

18 .449 1.283 85.702

19 .439 1.256 86.957

20 .404 1.154 88.112

21 .390 1.114 89.226

22 .371 1.060 90.286

23 .356 1.016 91.302

24 .343 .979 92.281

25 .333 .953 93.233

26 .305 .873 94.106

27 .296 .847 94.953

28 .276 .788 95.741

29 .250 .713 96.455

30 .243 .695 97.150

31 .233 .665 97.814

32 .228 .652 98.466

33 .198 .566 99.033

34 .191 .545 99.578

35 .148 .422 100.000

Extraction Method: Maximum Likelihood.

The rotated pattern matrix for the 35 items of the South African Psychological

Ownership Questionnaire (SAPOS) is displayed in Table 4.14. All the items had factor

loadings of .40 and higher, indicating the significance of these items for interpretative

purposes. Sixteen items loaded on factor 1, eight items loaded on factor 2, five items on

factor 3 and six items loaded on factor 4. The factors were labelled according to the

general content of their significant related items. The four factors of POSQ were labelled

Identification, Responsibility, Autonomy and Territoriality respectively.



Table 4.14: Rotated pattern matrix for the four-factor model

Factor

1 2 3 4

Q52 B .919 -.011 -.071 -.011

Q43 B .837 -.033 .047 -.019

Q51 SI .752 -.039 -.136 .035

Q31 B .742 .009 .175 -.066

Q56 SI .714 .056 .060 -.157

Q24 SI .704 -.058 .190 -.055

Q34 B .703 .018 -.201 -.161

Q66 SI .693 .006 .168 -.013

Q27 B .642 -.043 .052 .027

Q40 B .624 .098 .181 -.078

Q55 B .613 .003 .231 .003

Q12 SI .586 -.020 .127 -.023

Q49 B .551 .056 .036 .053

Q6 SI .547 -.028 086 .086

Q61 SI .539 .159 .154 .154

Q9 SI .456 .150 .002 .002

Q47 R .037 .795 .004 .081

Q54 R .071 .745 -.043 -.046

Q63 R .069 .706 -.071 .002

Q48 R .025 .678 .070 .064

Q62 R .017 .653 .008 .008

Q16 Acc -.057 .632 .096 -.006

Q59 R -.032 .630 -.057 -.081

Q28 Acc -.051 .558 .125 -.019

Q23Aut -.028 -.018 .775 .040

Q42 Aut .093 .064 .725 .038

Q29 Aut .108 .008 .705 -.036

Q19 Aut .014 .113 .689 -.079

Q38 Aut .217 -.012 .616 .113

Q11 Aut .074 .065 .598 .074

Q39 T -.063 .028 .032 .792

Q26 T -.125 -.009 .104 .700

Q35 T .035 .021 .014 .678

Q22 T .031 -.004 .045 .584

Q2 T .077 -.052 -.081 .470

Extraction Method: Maximum Likelihood. Rotation Method: Oblimin with Kaiser Normalization. a. Rotation converged in 6 iterations.



The inter-correlation matrix displayed in Table 4.15 shows the correlation coefficients

between the factors. Factor 4 (Territoriality) has little or no relationship with any other

factors because the correlation coefficients are very low (Field, 2005). However, factors

1 (Identification) and 2 (Responsibility), with an R = .363, and factors 1 (Identification)

and 3 (Autonomy) with an R = .466, correlated with one another, indicating that these

constructs are interrelated (Hair et al., 2006).

Table 4.15: Scale inter-correlation matrix for the four-factor model

Factor 1: Identification

Factor 2: Responsibility

Factor 3: Autonomy

Factor 4: Territoriality


1 1.000

Factor 2: Responsibility

2 .363 1.000

Factor 3: Autonomy

3 .466 .256 1.000


4 .105 -.048 -.047 1.000

A second-order factor analysis was performed using maximum-likelihood factor

extraction with an oblique rotation. The results of the second-order factor analysis

displayed in Table 4.16 indicate clearly the existence of two distinctive dimensions.

Factors 1, 2 and 3 (Identification, Responsibility and Autonomy) share common

variance and contribute significantly (R= .821, .683 and .767) to a single overall

dimension labelled promotive (promotion-orientated) psychological ownership. Factor 4

(Territoriality) loaded to the second single overall dimension (R = .984), labelled

preventative (prevention-orientated) psychological ownership.

Table 4.16: Rotated second-order factors from the matrix of factor correlations

2nd Order

1

2nd Order

2

Factor 1: Identification 1 .821 .218

Factor 2: Responsibility 2 .683 -.122

Factor 3: Autonomy 3 .767 -.053

Factor 4: Territoriality 4 -.012 .984



4.2.6.2 Examination of construct equivalence

Construct equivalence of the SAPOS were determined by using exploratory factor

analysis and target (Procrustean) rotation. The factor loadings of the different race

groups were rotated to a joint common matrix of factor loadings. After target rotation

had been carried out, factorial agreement was estimated using Tucker’s coefficient of

agreement (Tucker’s phi). However, due to the small representation of the Indian

(4.80%) and Coloured (3.40%) respondents construct equivalence could not be

determined for all cultural groups. These two race groups have been incorporated with

the African group based on the Employee Equity Act, 1997 (Act No. 75 of 1997) who

defined black people as a generic term for Africans, Coloureds and Indians. The

Tucker’s phi-coefficients for the two groups are given in Table 4.17.

Table 4.17: Construct equivalence of the SAPOS for the two groups

Group Percentage of

sample

Tucker’s phi –

Identity

Tucker’s phi -

Responsibility

Tucker’s phi -

Autonomy

Tucker’s phi -

Territoriality

White 59.54 .93 .96 .94 .96

Black 40.46 .93 .96 .94 .96

Inspection of Table 4.17 shows that the Tucker’s phi coefficients for the Black and White

respondents were all acceptable (>.90). Therefore, it can be deduced that the four

factors of the SAPOS were equivalent for the two race groups.

4.2.6.3 Examination of internal consistency

Evidence of internal consistency could be provided by a number of measures. In this

study reliability was calculated using Cronbach’s alpha and variance extracted (VE)

estimates. Alphas for each subscale were highly satisfactory, ranging between .78 and

.94 and well above the .7 cut-off (Hair et al., 2006; Cortina, 1993). The variance

extracted (VE) estimate is the average squared factor loading and, according to Hair et

al. (2006), as a rule of thumb a VE value of .50 or higher indicates adequate



convergence. Alpha coefficients and variance extracted (VE) estimates for the

subscales are displayed in Table 4.18.

Table 4.18: Internal consistency for the subscales of SAPOS Corrected Item-Total

Correlation Cronbach's Alpha if

Item Deleted


Q6 SI .497 .940 Q9 SI .494 .939 Q12 SI .636 .936 Q24 SI .757 .933 Q27 B .628 .937 Q31 B .795 .932 Q34 B .769 .933 Q40 B .714 .932 Q43 B .818 .932 Q49 B .579 .938 Q51 SI .633 .937 Q52 B .844 .931 Q55 B .699 .935 Q56 SI .724 .934 Q61 SI .507 .939 Q66 SI .763 .933 Scale reliability: .939 Variance Extracted (VE): .460 Factor 2: Responsibility Q16 Acc .602 .858 Q28 Acc .523 .866 Q47 R .738 .845 Q48 R .634 .855 Q54 R .716 .846 Q59 R .567 .863 Q62 R .598 .859 Q63 R .664 .852 Scale reliability: .871

Variance Extracted (VE): .460

Factor 3: Autonomy

Q11 Aut .631 .862 Q19 Aut .654 .857 Q23 Aut .673 .854 Q29 Aut .681 .853 Q38 Aut .691 .853 Q42 Aut .764 .838



Although the alphas for the subscales were highly satisfactory, the variance extracted

(VE) estimates of .460 (Identification); .460 (Responsibility); .470 for Autonomy and .430

for Territoriality were less than .50. According to Hair et al. (2006), this could be an

indication that variance due to measurement error is larger than the variance depicted

by the factor. However, Hatcher (1994) notes that the variance extracted estimate test is

very conservative, therefore reliabilities can be acceptable even if variance extracted

estimates are less than .50.

4.2.6.4 Descriptive statistics of the scales of the SAPOS

The results of the descriptive statistics of the South African Psychological Ownership

Questionnaire (SAPOS) for the four factors are set out in Table 4.19: The mean,

standard deviation, skewness and kurtosis were computed for the sample scores on the

four scales of the SAPOS.

Scale reliability: .874



Q2 T .428 .778 Q22 T .552 .734 Q26 T .536 .740 Q35 T .599 .718 Q39 T .649 .703 Scale reliability : .776




Table 4.19: Descriptive statistics for the four scales of the SAPOS

Factor 1:

Identity

Factor 2:

Responsibility

Factor 3:

Autonomy

Factor 4:

Territoriality

N 356 356 356 356

Mean 72.3258 41.3146 27.8258 17.6601

Std. Error of Mean .69667 .22004 .27617 .27821

Std. Deviation 13.14472 4.15162 5.21082 5.24925

Skewness -1.126 -.930 -1.189 .134

Skewness error .129 .129 .129 129

Kurtosis 1.534 2.827 2.031 -.725

Kurtosis error .258 .258 .258 .258

According to Field (2005), the values of skewness and kurtosis are 0 within a normal

distribution. Therefore, values of skewness or kurtosis above or below 0 indicate a

deviation from normal. Morgan and Griego (1998), on the other hand, state that the

assumption for normality expects skewness and kurtosis to be less than 2.5 times the

standard error. According to these criteria the summated scores of the sample on the

four scales presented in Table 4.19 indicate that the data has a deviation from the

normal distribution with a tendency towards negative skewness and leptokurtic

distributions.

4.2.7 Step 7: Confirmatory factor analysis

Following the guideline of Krzysofiak et al. (1988), the original sample was randomly

split into two halves. One half of the sample (n = 365) was used for the development of

a model (as discussed in Chapter 3) and the other half was used to validate the

outcome that was obtained from the first half (Gerbing & Anderson, 1988). Therefore,

the 35-item psychological ownership measure (SAPOS) was subjected to confirmatory

factor analysis (CFA) on the other half of the sample (n = 365) and to competing model

comparisons using EQS. Conducting a CFA allowed the researcher to examine the

dimensionality and to evaluate the internal consistency of the developed measure more

rigorously (Faranda, 2001).



The data was checked for the presence of multivariate outliers because these outliers

might unduly influence the results of the factor analysis (Tabachnick & Fidell, 2007)

Multivariate outliers were identified by inspecting the standardised scores (z-scores)

and Mahalanobis distance statistic. There were no cases with very large Mahalanobis

distance values that were clearly separated from the values of the other cases.

Assumptions of normality were assessed as well. According to Bentler (as cited in

Byrne, 2006) Mardia’s normalised estimate values greater than 5.00 indicate that the

data are non-normally distributed. In this study, Mardia’s coefficient (397.433) and the

normalised estimate of the coefficient (z-statistic) of 85.6875 suggested that the

measured variables were not normally distributed. Tabachnick and Fidell (2007)

suggest that in the case of non-normality the robust maximum-likelihood (ML) estimate

with the Santorra-Bentler scaled chi-square and adjustment of the standard errors

should be employed. According to Garson (2002), the Satorra-Bentler chi-square is a

corrected chi-square that makes an attempt to rectify the bias that is presented when

the data are noticeably non-normal in the distribution.

The structural equation models for the four dimensions underlying the SAPOS are

depicted in Table 4.20 and Figure 4.5 respectively. Latent variables were allowed to

correlate with one another.

Table 4.20: Maximum-likelihood estimates of the four-factor model (n = 365)

Fit indices Four-factor solution

S-B χ2 951.772

Df 554

NNFI 0.897

CFI 0.904

RMSEA 0.045 (0.04 – 0.05)

SRMR 0.059

The result of the Satorra-Bentler chi-square statistic was 951.772, based upon 554 df (p

< .0001). This chi-square statistic is significant and revealed a poor overall fit of the



original measured four-factor SAPOS model. If a chi-square value is significant, it

indicates the covariance structure of the model differs significantly from the observed

covariance structure. A non-significant chi-square value indicates a good model fit

(Garson, 2002). However, according to Kelloway (1998), given the sample size and chi-

square/df ratio, it would be incorrect to accept a poor model fit based on the significance

of the chi-square index alone. The chi-square/df ratio was 1.72. Ullman (as cited in

Tabachnick & Fidell, 2007) suggests that ratios of two or less can be interpreted as an

indication of a good fit. According to this guideline, the measurement model appears to

fit the data well. However, the chi-square statistic is very sensitive to sample size

(Garson, 2002), with the result that a model with a large chi-square may still have a

good fit if the fit indices are high. Therefore, the chi-square statistic must be used with

caution and other multiple fit indices should be used to assess a model’s goodness-of-

fit. According to Bentler (2007) standardised root mean square residual (SRMR) should

be reported, accompanied by at most two other fit indices of fit, such as the comparative

fit index (CFI).

The model yielded a CFI value of .904. This value is slightly greater than the required

.90 but less than the more recently .95 desirable levels (Hu & Bentler, 1999) to indicate

a good model fit.

The RMSEA value was estimated at .045. This RMSEA value supports the belief of a

good model fit because, according to Hair et al. (2006), RMSEA values between .05

and .08 are indicative of an acceptable fit. In a well-fitting model, the 90% confidence

interval of the RMSEA should be between 0 and .08. The 90% confidence interval of the

RMSEA (.04, .05) confirmed the acceptable fit of the four-factor measurement model to

the data.

The model yielded an SRMR value of .059. Considering the guideline of Garson (2002)

that SRMR values of less than .05 are widely considered good fit, and below .08

adequate fit, this value illustrates a fairly good fit.



The chi-square/df ratio, CFI, RMSEA and SRMR values therefore met the minimum

recommended standards, indicating a reasonable model fit.

The path diagram with parameter estimates produced by EQS, based on the four-factor

results of the EFA, is displayed in Figure 4.5. The observed variables in this diagram

reflecting the 35 items of the POSQ are coloured in blue, pink, green and purple,

whereas the independent variables, namely Identification, Responsibility, Autonomy and

Territoriality, are coloured in yellow. The 35 one-way arrows are indicative of regression

coefficients that are indicative of the hypothesised effects of the observed variables

(Bentler, 2004), whereas the two-way arrows represent the correlation or covariance

between variables.

Path values indicate measures of reliability, displaying how well the observed variable

explains each latent construct. Values of .7 or higher are required for acceptance in this

regard (Hair et al., 2006).

For the latent construct Identification (F1) the observed variables Q6, Q9, Q12, Q49,

Q51, Q61 and Q66 demonstrated moderate to low path values varying between .45 and

.65, which is as a rule not considered as a good coefficient for acceptability. These

observed variables also show high standard error, indicating that they cannot be

sufficiently explained by the latent construct Identification. This finding is indicative of a

limited fit with judgement or sound theory regarding the relationship between the latent

construct and these observed variables. Except for variable Q49, the other observed

variables had originally been defined to form part of the self-identity dimension.



Figure 4.5: Standardised estimated parameters of the four-factor model



However, on the basis of the researcher’s belief at that point, she decided to retain

these items as part of one dimension because these constructs seemed to be very

similar; according to Lee (as cited in Edwards, 2005, p. 210), identification involves a

sense of belongingness, and this is supported by Ashforth and Mael (1989, p. 21), who

referred to identification as “the perception of oneness or belongingness”.

Identification is a very complex phenomenon (Edwards & Peccei, 2007). Edwards and

Peccei developed an instrument for measuring organisational identification (OID).

According to them OID, is a multi-dimensional construct comprising three empirically

distinct, yet strongly related components: self-categorisation and labelling; value and

goal synergy; and belonging and membership. “Self-categorisation and labelling” refers

to the process by which individuals categorise themselves as members of the

organisation as a social category and, through the process, effectively label themselves

as organisational members (Ashforth & Humphrey, as cited in Edwards & Peccei

(2007). “Value and goal synergy” refers to the extent to which employees share the

values and goals of the organisation and integrate them into their own belief system

(Ashforth & Mael, 1989). “Belonging and membership” refers to the extent to which

employees experience a sense of attachment to, belonging and membership of the

organisation (Brown, 1969; Lee, as cited in Edwards, 2005). Edwards and Peccei found

that although these three subcomponents are analytically and empirically distinct, they

were found to be strongly interrelated and therefore showed very low discriminant

validity.

According to Edwards and Peccei (2007), in practice it might not be meaningful and/or

sensible to treat the three subcomponents as completely separate constructs and use

them as distinct variables in analysis. They therefore suggest that it may rather be

preferable to combine the three subcomponents into an aggregate measure and use it

as a single overall scale of OID.

Avey et al. (2009) argue that although the underlying principle of ownership may be

manifested in both self-identity and belongingness, these two should remain distinct, yet



related constructs. However, preliminary exploratory factor results conducted on the

psychological ownership questionnaire (POQ) developed by Avey et al. (2009) on a

South African sample comprising 145 health professional employees also indicated that

the self-identity and sense of belonging items load onto one factor.

Higher, more acceptable path values were displayed by the remaining observed

variables, ranging between .70 and .85.

In the case of Responsibility (F2), the highest coefficient is displayed by variable Q47

(.80) and the lowest coefficient by variable Q59 (.59).

In the case of Autonomy (F3), an overall better indication of reliability in comparison with

the previous latent constructs is displayed, with the lowest path value of .63 for variable

Q11.The other coefficients vary from .68 to .84.

Territoriality (F4) displays coefficients ranging between .48 and .81. Variable Q2 has the

lowest path value (.48) and variable Q39 the highest path value (.81). Once again, in

the case of variable Q2, it seems that the logically and theoretically aligned concept

cannot be adequately explained by the latent construct.

The latent constructs were also allowed to correlate. Theoretically, well-defined

constructs should reveal low values. The correlations in this model between the latent

constructs Identity and Responsibility and Responsibility and Autonomy were relatively

low, with values of .46 and .48 respectively. A moderate correlation with the value of .65

was found between Identity and Autonomy. As expected, and confirmed by the second-

order factor analysis, the latent construct, Territoriality (the preventative form of

psychological ownership) showed either no (.00) or extremely low negative correlations

(-.09 and -.03) with the other more promotive latent constructs.



4.2.8 Step 8: Discriminant and criterion-related validity

4.2.8.1 Discriminant validity

Discriminant validity is the extent to which a construct is truly distinct from other

constructs (Fornell & Larcker, 1981). In this study, evidence of discriminant validity was

provided by the following procedure recommended by Fornell & Larcker. The variance

extracted (VE) estimates (the average squared factor loading (Hair et al., (2006)) for two

factors were compared and then compared with the squared correlation between the

two factors. If the variance extracted estimates for both factors exceeds the squared

correlation, discriminant validity is demonstrated. The correlations and squared

correlations between the factors are depicted in Table 4.21.

Table 4.21: Correlations and squared correlations between the four factors

Factor 1:

Identification

Factor 2:

Responsibility

Factor 3:

Autonomy

Factor 4:

Territoriality

Factor 1:

Identification

1 1.000 (.132) (.217) (.011)

Factor 2:

Responsibility

2 .363 1.000 (.067) (.002)

Factor 3:

Autonomy

3 .466 .256 1.000 (.002)

Factor 4:

Territoriality

4 .105 -.048 -.047 1.000

Variance

Extracted (VE)

.460 .460 .470 .430

Note: Values shown in brackets above the diagonal are squared correlations

As illustrated in Table 4.21, the correlation between factor 1 (Identification) and factor 2

(Responsibility) for example is .363 and the squared correlation is .132. The variance

extracted estimate for both factor 1 (Identification) and factor 2 (Responsibility) is .460.

The discriminant validity of factor 1 (Identification) and factor 2 (Responsibility) was



confirmed, because the variance extracted estimates exceeded the square of the inter-

factor correlation.

The correlation between factor 2 (Responsibility) and factor 3 (Autonomy) is .256 and

the squared correlation is .067. The variance extracted estimates for factor 2

(Responsibility) and factor 3 (Autonomy) are .460 and .470 respectively. Since the

variance extracted estimates exceeded the square of the inter-factor correlation, the

discriminant validity of factor 2 (Responsibility) and factor 3 (Autonomy) is confirmed.

The correlation between factor 1 (Identification) and factor 3 (Autonomy) is .466 and the

squared correlation is .217. The variance extracted estimates for factor 1 (Identification)

and factor 3 (Autonomy) are .460 and .470 respectively, therefore, similarly, the

discriminant validity of factor 1 (Identification) and factor 3 (Autonomy) is confirmed, due

to the fact that the variance extracted estimates exceeded the square of the inter-factor

correlation.

In the case of factor 4 (Territoriality), examination of the other variance extracted

estimates (.460, .460 and .470) and squared correlation coefficients (.011, .002 and

.002) confirmed discriminate validity within the model.

4.2.8.2 Criterion-related validity

According to Cronbach and Meehl (1955), the relationship of the focal construct with

other similar constructs should be examined to develop a nomological network. In this

study, evidence of criterion-related validity was provided by examining the relationships

between measures of psychological ownership and other theoretically related

constructs, such as organisational commitment (Avey et al., 2009; Mayhew et al., 2007;

O’Driscoll et al, 2006) Other measures that were included in the primary study to

explore the corresponding semantic network were job satisfaction and turnover

intentions (Avey et al., 2009; Buchko, 1993; Mayhew et al., 2007; Pierce et al., 1991;

VandeWalle et al., 1995; Van Dyne & Pierce, 2004). According to Bagozzi and Yi



(1988), it is essential that researchers carefully scrutinise the internal structure of their

model even if global measures of fit imply a satisfactory model. Therefore, for the

purposes of this study, the composite reliability of promotive psychological ownership

was calculated, as suggested by Schepers (1992). A high composite reliability value of

.945 was calculated for the promotive psychological ownership dimension. According to

Bagozzi and Yi, composite reliability values of greater than .60 are desirable.

• Organisational commitment

To assess organisational commitment, all eight items from Allen and Meyer’s (1991)

Affective Organisational Commitment Scale were used. Affective commitment refers to

“the employee’s emotional attachment to, identification with, and involvement in the

organization” and “[E]mployees with a strong affective commitment continue

employment with the organization because they want to do so” (Meyer & Allen, 1991, p.

67). Both VandeWalle et al. (1995) and Van Dyne and Pierce (2004) have shown that

affective commitment is related to psychological ownership. The eight items were

measured on a seven-point scale, ranging from 1 = strongly disagree to 7 = strongly

agree. Allen and Meyer (1990) found a coefficient alpha of .87 for the affective

commitment scale. In this study, the affective commitment scale yielded an acceptable

reliability alpha (� = .71).

• Job satisfaction

Job satisfaction was measured using three items that form part of Hackman and

Oldham’s (1980) Job Diagnostic Survey. Although Idaszak et al. (1988) suggest that an

instrument should have at least four to six items per scale, Tabachnick and Fidell (2007)

suggest retaining at least three items per factor. According to Tabachnick and Fidell

(2007), it is possible to retain a factor with only two items if the items are highly

correlated (i.e., r > .70) and relatively uncorrelated with other variables. The three items

were measured on a seven-point scale, ranging from 1 = disagree strongly to 7 = agree

strongly. This scale demonstrated internal reliability of .65.



• Turnover intentions

Turnover intentions were assessed by using a three-item turnover intention scale used

by O’Driscoll and Beehr (1994). Zwick and Velicer (1986) and Gorsuch (1997) suggest

that factors need at least three to four substantial factor loadings within the .40 (after

rotation) range to be considered substantially meaningful. The three items rated

whether respondents thought about leaving their job, planned to look for a new job over

the next 12 months, and would actively search for a new job outside the organisation.

O’Driscoll and Beehr (1994) obtained an alpha of .93 from the three-item turnover

intention scale in their study@�The three items were measured on a six-point rating scale

where the response format varied for each item. The Cronbach alpha for turnover

intentions was .77.

Correlation results

Correlation between the constructs was determined by means of Pearson correlation.

Although the distribution is skewed, it was more useful to employ Pearson correlation

because of the relatively large sample size (N = 713). As opposed to the more

promotion-oriented forms of psychological ownership, feelings of territoriality showed no

relationship with the outcome variables. Pierce et al. (1991) propose that psychological

ownership is an antecedent to organisational commitment. As anticipated, and in

accordance with empirical research findings by Avey et al. (2009), Mayhew et al. (2007)

and O’Driscoll et al. (2006), promotive psychological ownership was positively related to

affective commitment toward the organisation with an r = .642 (p < 0.01). A positive

relationship was confirmed between job satisfaction and promotive psychological

ownership (r = .536, p < 0.01). Empirical research findings (e.g. Avey et al. (2009),

Buchko (1993), Mayhew et al. (2007), Pierce et al. (2007), VandeWalle et al. (1995) and

Van Dyne and Pierce (2004)) support a positive relationship between job satisfaction

and psychological ownership. According to Mayhew et al., although evidence indicates

relationships between psychological ownership and commitment and job satisfaction,

the emergence of such relationships may be a consequence of conceptual overlap.



The distinctiveness of the psychological ownership construct was supported by previous

research done by Van Dyne and Pierce. They found that organisation-based

psychological ownership could clearly be distinguished from both affective commitment

and job satisfaction.

As expected and confirmed by Avey et al. (2009), turnover intentions were negatively

related to promotive psychological ownership with an r = - .376 (p < 0.01).

The correlation results for all study variables are reported in Table 4.22.



Table 4.22: Relationships with Psychological Ownership

Promotive psychological

ownership

Identification Responsibility Autonomy Territoriality Commitment Job satisfaction

Turnover intentions

Promotive psychological ownership

Pearson Correlation

1 .944** .614** .757** .011 .642** .536** -.376**

Sig. (2-tailed) .000 .000 .000 .759 .000 .000 .000

N 713 713 713 713 713 713 708 709 Identification Pearson

Correlation .944** 1 .416** .575** .040 .675** .526** -.412**

Sig. (2-tailed) .000 .000 .000 .286 .000 .000 .000

N 713 713 713 713 713 713 708 709 Responsibility Pearson

Correlation .614** .416** 1 .364** -.030 .324** .249** -.115**

Sig. (2-tailed) .000 .000 .000 .417 .000 .000 .002

N 713 713 713 713 713 713 708 709 Autonomy Pearson

Correlation .757** .575** .364** 1 -.033 .371** .420** -.236**

Sig. (2-tailed) .000 .000 .000 .375 .000 .000 .000

N 713 713 713 713 713 713 708 709 Territoriality Pearson

Correlation .011 .040 -.030 -.033 1 -.071 -.147** .156**

Sig. (2-tailed) .759 .286 .417 .375 .058 .000 .000

N 713 713 713 713 713 713 708 709 Commitment Pearson

Correlation .642** .675** .324** .371** -.071 1 .467** -.459**

Sig. (2-tailed) .000 .000 .000 .000 .058 .000 .000

N 713 713 713 713 713 713 708 709 Job satisfaction Pearson

Correlation .536** .526** .249** .420** -.147** .467** 1 -.475**

Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000

N 708 708 708 708 708 708 708 708 Turnover intentions

Pearson Correlation

-.376** -.412** -.115** -.236** .156** -.459** -.475** 1

Sig. (2-tailed) .000 .000 .002 .000 .000 .000 .000 N 709 709 709 709 709 709 708 709

**. Correlation is significant at the .01 level (2-tailed).



4.2.8.3 Comparing different groups

Independent sample t-tests and the analysis of variance (ANOVA) technique were

conducted to assess whether employees varying in biographical variables (age, gender,

ethnic group, education, the sector in which their organisation operates, level in the

organisation and registration with a professional board) differed significantly with regard

to the specific dimensions (Identification, Responsibility, Autonomy and Territoriality)

underlying the concept of psychological ownership. Independent sample t-tests were

used to test whether significant differences existed between the means of two groups,

and where several independent variables were compared the analysis of variance

(ANOVA) technique was used.

Independent sample t-tests were conducted for the following biographical variables:

• Ethnic group

• Gender

• Registration at a professional board

ANOVAs were conducted for the following biographical variables:

• Level in the organisation

• Age

• Educational level

• Sector in which the organisation operates

• Ethnic groups

According to Pierce et al. (2003), cultural aspects of a social context may have a

significant influence on people’s psychological ownership. Pierce and colleagues

suggest that it is possible that feelings of ownership may be present in different cultures

to a different extent.



According to Pierce et al. (2003), different targets and different characteristics of

potential targets of psychological ownership may become more important in different

cultures. The different types of target towards which people develop feelings of

ownership will depend on where the self-concept of individuals in a given society

predominantly resides. For example, cultures that are more collectivistic, familial and

relationship based tend to be more orientated towards family and friends, while cultures

that are more individualistic gain their self-concept primarily from their personal

successes and achievements. It can be expected that the former will develop feelings of

ownership primarily towards social targets like people and family, while the latter would

focus more on their material possessions and work that address these achievements.

According to Maré (2009, p 84),

Even without sound theoretical analysis it is evident that the white South

African is strongly representative of an independent individualist style.

Black South Africans, on the other hand, constitute a cohesive approach

that can be deemed rather the opposite of individualism.

According to Arnoldi-Van der Walt (as cited in Maré, 2009), a contrast exists between

the individualistic understanding which is manifested in the western model of society

and the ubuntu model. Ubuntu is an African humanist philosophy that focuses on

people’s interconnectedness with one another. According to the Ubuntu belief, a person

cannot exist as a human being in isolation because your actions affect the whole world.

Arnoldi-Van der Walt (as cited in Maré, 2009, p 87) compares the Western and Ubuntu

models with regard to the social/environmental and cultural context, as presented in

Table 4.23.



Table 4.23: Differences in social/environmental/cultural context between Western and Ubuntu models

Western model Ubuntu model

Individualistic culture: view themselves as individuals who together form a group; self reliant

High collectivist culture: group affiliation much more important than individuality; dependent

Low face-saving culture: very direct High face-saving culture: dignity; inclusive group orientation; greatest fear is rejection by the group

Low-context culture: information explicitly conveyed High-context culture: information implicitly conveyed; unconditional dialogue

In the forefront of development/utilization of modern technology

Low utilization of management of foreign technology

The more the individual has, the more powerful he/she is

The more the communal person is prepared to give and share, the more respected he/she is

Source: Maré (2009, p. 87)

An independent t-test was used to determine the difference between black and white

South African employees with regard to the dimensions of psychological ownership.

Differences in the mean scores would indicate that the two groups differ with regard to

their perception of psychological ownership. The results of the independent t-test are

displayed in Table 4.24. According to Levene’s test, this study found that a significant (p

< .05) difference exists between black employees and white employees, specifically

towards the Responsibility and Territoriality dimensions of psychological ownership.

With regard to the Responsibility dimension of psychological ownership, white

employees (mean = 41.907) showed a slightly higher inclination towards responsibility

in their organisations than black employees (mean = 40.328). This indicates that white

employees still take full responsibility for the organisations in which they are currently

employed in order to secure their jobs. They cannot afford to loose their jobs since

competition in the market is stronger than ever due to the implementation of

employment equity (EE) and black economic empowerment (BEE). This is supported by

a study conducted by Berg, Buys, Schaap and Olckers (2004) on Scheper’s Locus of

Control Inventory. They found differences between English second-language (black

people) and first-language (white people) groups with regard to their internal locus of

control (the individual believes that outcomes are a consequence of her or his own

behaviour). The first-language group showed a higher internal locus of control than the



second-language group. The practical significance (d = .355) between black employees

and white employees with regard to responsibility reflects a medium effect size.

The difference that exists between black employees (mean = 19.729) and white

employees (mean = 16.452) with regard to Territoriality can probably be ascribed to

black people’s collectivistic culture. This is related to a study conducted by Watkins

(1995) among 487 managers working in various South African organisations, which

found that black managers experience a stronger sense of entitlement than their white

colleagues, which is ascribed to deprivation in the past. The difference between black

employees and white employees with regard to territoriality reflects a medium effect size

(d = .658).

Table 4.24: t-test: Ethnic groups

Dependent

variables

Groups N Mean Standard

Deviation

t-value Sig.

(2-tailed)

d

Identification White 418 71.794 13.268 -1.717 .086 .132

Black 284 73.514 12.665

Responsibility White 418 41.907 3.509 4.610 .000* .355

Black 284 40.328 4.998

Autonomy White 418 27.845 5.313 .443 .658 .034

Black 284 27.666 5.178

Territoriality White 418 16.452 4.812 -8.615 .000* .658

Black 284 19.729 5.137

E(F5�>

• Gender

Results of the independent t-test as displayed in Table 4.25 indicated that a significant

difference exists between males and females with regard to the Responsibility and

Territoriality dimensions. Males (mean = 41.760), on the one hand, seem to feel more

responsible and take more responsibility than their female counterparts mean = 40.886),

whereas females (mean = 18.571), on the other hand, tend to be more territorial than



males (mean = 16.728). The reason for this might be the fact that the workplace is

traditionally occupied by males and thus it is an environment in which males take up a

natural role and the accountability for that role. One may say that males are groomed

for these roles through the cultural view that the male is expected to be the

breadwinner. This is supported by Coetzee (2008), who found in her study that males

and females differed significantly with respect to their career preferences. The male

respondents showed significantly higher mean scores than the females with regard to

managerial (referring to individuals who view upward mobility to positions of

successively greater responsibility) and autonomy (referring to independence and

freedom from external interruptions) career preferences.

According to Lamphere (1985), our culture still assumes that women are primarily

daughters, wives, and mothers and therefore are expected to care for the children and

household, thus their achievements in the workplace involve a much more arduous

journey; they have to fight for their equivalent place in the corporate world and this is

evident in their higher territorial behaviour. Although significant differences were found

between males and females with regard to these two dimensions, a medium effect size

(d = .209 and d = .358) was calculated.

Table 4.25: t-test: Gender

Dependent

variables


Deviation

t-value Sig.

(2-tailed)

d

Identification Male 287 72.265 13.376 -.415 .678 .032

Female 420 72.679 12.749

Responsibility Male 287 41.76 3.958 2.699 .007* .209

Female 420 40.886 4.402

Autonomy Male 287 27.728 5.097 -.161 .872 .012

Female 420 27.793 5.346

Territoriality Male 287 16.728 5.116 -4.681 .000* .358

Female 420 18.571 5.158

E(F5�>



• Registration with a professional board

An independent t-test was conducted between employees registered with a professional

board and those that were not registered. Table 4.26 indicates a significant difference

between employees registered with a professional board and those not registered with

regard to the Identification and Territoriality dimensions. People who are registered

(mean = 73.605) seem to have a higher feeling of identification than those who are not

registered (mean = 18.784), as well as a greater territorial need than those who are not

registered (mean = 71.395 and mean = 16.678 respectively). Since a professional

qualification indicates that the person is a competent professional, whose skills and

knowledge are highly valued by industry peers, a professional qualification allows

employees to plan their continuous professional development (CPD) with structure and

direction that can lead to the enhancement of their career prospects; provision of higher

earning potential and greater status and influence (Institute of Telecommunications

Professionals).

A significant portion of the research population consisted of professionals working in the

mining and engineering field, where there are stringent regulatory obligations in their

professional environment. Stringent laws may result in non-compliant professionals

receiving harsh penalties or even jail sentences. The result of this can be seen in their

higher mean with regard to Identification and Territoriality.

The practical significance value for Identification was d = .169, reflecting a small effect

size and therefore negligible in terms of practical importance. The practical significance

(d = .412) for Territoriality, however, reflected a medium effect size.



Table 4.26: t-test: Registration with a professional board

Dependent

variables


Deviation

t-value Sig.

(2-tailed)

d

Identification Registered 324 73.605 12.870 2.205 .028* .169

Not-registered 357 71.395 13.235

Responsibility Registered 324 41.043 4.771 -1.717 .086 .133


Autonomy Registered 324 28.034 4.821 1.039 .299 .08


Territoriality Registered 324 18.784 5.194 5.383 .000* .412


E(F5�>

• Age

The ANOVA results displayed in Table 4.27 indicate that there is a significant (F =

6.302; p < .05) difference between the age groups with regard to their territorial need. A

Sheffé post-hoc test was conducted to determine exactly where the difference lies.

Employees in the age category 40–49 years (mean = 19.00) seem to be more territorial

than those under 29 years of age (mean = 17.285), those between 30 and 39 years of

age (mean = 17.291), and those older than 50 years of age (mean = 17.083).

Employees older than 50 are less territorial since most organisations today allow staff to

take early retirement at age 55. The respondents who are on average older than 50 (n =

123) are therefore less territorial, since they have an exit option out of the organisation

by requesting early retirement. This does not, however, apply to the age group 40–49

years, who in the current economic downturn are at high risk of retrenchment and a

probable target of transformation. Their territorial score is much higher, confirming their

fight for corporate survival, while employees younger than 40 years in generation X are

still confident that they can climb the corporate ladder either in their current or an

alternative organisation. This is confirmed by a study conducted by Coetzee (2008),

who found that employees in the middle and late adulthood life stages showed a higher



need for developing expertise by means of further growth and learning opportunities

than those in the early adulthood life stage. This confirms their belief that with the

acquired skills they will improve their employability and will be able to prove their ability

to their superiors.

The effect size was partial eta2 with a value of .027, which is an indication of a small

effect size. Although the partial eta2 value was small, determining effect sizes between

specific groups reveals medium effect sizes as follows: a d = .322 value between the

group under 29 years of age and those between 40 and 49 years of age, a value of d =

.342 between 30–39 years and 40–49 years old, and a d = .352 value between those

employees between 40 and 49 and those older than 50 years.

Table 4.27: ANOVA and post-hoc Scheffé’s test – Age

Dependent variables

F – value Sig. Sub-groups

N Mean Standard Deviation

Partial eta squared

Identification 2.331 .073 <29 137 70.839 13.516 .010

30-39 213 71.667 12.187

40-49 240 73.163 13.167

50+ 122 74.602 13.208

Responsibility 1.354 .256 <29 137 40.883 3.963 .006

30-39 213 41.366 4.195

40-49 240 41.042 4.280

50+ 122 41.829 4.517

Autonomy 1.420 .236 <29 137 27.314 4.940 .006

30-39 213 27.418 5.715

40-49 240 28.092 4.889

50+ 122 28.317 5.280

Territoriality 6.302 .000* <29 137 17.285a 5.229 .027

30-39 213 17.291b 4.567

40-49 240 19.00abc 5.386

50+ 122 17.083c 5.492

*p< .05



• Educational level

Using the Scheffé’s test of post-hoc comparisons, the following groups differed

significantly on the Identification (F = 4.063; p < .05), Responsibility (F = 4.083; p < .05),

Autonomy (F = 3.000; p < .05), and Territoriality (F = 13.249; p < .05), dimensions:

employees with a bachelor’s degree (mean = 74.748) showed a stronger identification

with the organisation than employees with postgraduate degrees (mean = 70.543).

Postgraduates (mean = 41.732) showed a slightly higher inclination towards

responsibility in their organisations than employees having a diploma (mean = 40.493).

Given that postgraduates are probably employed in senior management positions, this

is another indication that, socially and by law, strenuous responsibility is placed on

senior management for decisions made and actions taken. According to Coetzee (2008,

p. 18), “people with a post-graduate qualification seem to have a higher sense of career

calling and are more interested in applying their expertise where they can help make a

difference in their own and other’s lives”.

Postgraduates (mean = 28.093) experience more autonomy in the workplace than

employees with only Grade 12 (mean = 25.883) as their highest qualification. This is

supported by Brass’s (1985) observation that employees exposed to high job design

autonomy experienced more influence (control) than their counterparts working with low

autonomy. This assumption is based on the fact that postgraduates are probably

employed in higher management positions where they have the freedom to schedule

work and determine how it is done (Ashforth & Saks, 2000)

With regard to the Territoriality dimension, employees with a diploma (mean = 19.331)

showed a higher territorial need than those with a bachelor’s degree (mean = 16.807) or

postgraduate degree (mean = 16.736). In a typical organisational hierarchy the number

of junior management employees competing for the next middle management position

is in all probability 10:1, while the contenders for the next senior management positions

are two or three candidates. This assumption is based on the fact that postgraduates

are employed at the senior management level. According to Coetzee and Schreuder (as



cited in Coetzee, 2008), people with an undergraduate qualification seem to be more

job orientated and are more attracted to careers that afford them the opportunity to

apply and develop their own skills in positions of power and influence.

Partial eta2 showed small effect size values (.17, .17, .13 and .54 respectively).

However, the calculation of effect sizes between the different groups showed the

following results: a medium effect size of d = .33 between those with a bachelor’s

degree and those with a postgraduate degree with regard to identification; a medium

effect size of d = .283 between those employees holding a diploma and those with a

postgraduate degree with regard to responsibility; a medium effect size of d = .397

between employees with a Grade 12 and employees with a postgraduate degree with

regard to the autonomy dimension. With regard to the territoriality dimension, a medium

effect size of d = .475 was calculated between diploma holders and bachelor degree

holders, as well as between diploma holders and postgraduate degree holders with a

value of d = .515. The results of the ANOVA are displayed in Table 4.28.



Table 4.28: ANOVA and post-hoc Scheffé’s test – Education

Dependent variables

F - value Sig. Sub-groups N Mean Standard Deviation

Partial eta squared

Identification 4.063 .007* Grade 12 60 72.1500 12.513 .017 Diploma 223 73.646 13.029 Bachelor’s

degree 135 74.748* 11.842

Post-graduate 280 70.543* 13.555 Responsibility 4.083 .007* Grade 12 60 41.383 3.532 .017 Diploma 223 40.493* 4.812 Bachelor’s

degree 135 41.674 4.055

Post-graduate 280 41.732* 3.879 Autonomy 3.000 .030* Grade 12 60 25.883* 5.666 .013 Diploma 223 27.803 5.000 Bachelor’s

degree 135 27.948 5.017

Post-graduate 280 28.093* 5.435 Territoriality 13.249 .000* Grade 12 60 18.667 5.401 .054 Diploma 223 19.331ab 5.426 Bachelor’s

degree 135 16.807a 5.165

Post-graduate 280 16.736b 4.604 *p < .05

• Sector in which organisation operates

ANOVA results displayed in Table 4.29 indicate a significant difference in the extent to

which the different sectors in which the organisation operates vary, with regard to the

Responsibility (F = 4.880; p < .05) and Territoriality (F = 16.084; p < .05) dimensions.

Once more the post-hoc test of Sheffé was employed to determine where the

differences between the groups lie. Employees working in the financial sector (mean =

2.67) seem to be more responsible than those employed in the government sector

(mean = 40.506), or those working in other sectors (mean = 40.919) such as

telecommunications, information technology, professional services and others.

Employees working in the government sector (mean = 19.913) seem to be more

territorial than those employed in the financial sector (mean = 16.301), chemical

industry (mean = 16.204) as well as manufacturing and production (mean = 17.182)



which include the mining sector. Those employees employed in other sectors (mean =

18.626) such as telecommunications, information technology and professional services

showed a higher territorial need than those working in either the financial sector (mean

= 16.301) or the chemical industry (mean = 16.204).

Financial service employees are subject to stringent regulatory compliance laws both in

South Africa and internationally (e.g. the BASIL–code). Employees within the financial

services organisations write annual compliance exams to verify their knowledge and

understanding of these laws and they are not allowed to perform their duties if these

exams are not completed in a specified period. The FAIS Act (Financial Advisory and

Intermediary Services Act), for example, was introduced to regulate all financial

advisors and intermediaries. The aim of this act is to protect the consumer against

inappropriate financial advice. The financial service sectors’ responsibility score is much

higher than that of other organisations, since the responsibility of passing these

compliance exams has been internalised and each employee is responsible for

performance in this regard in his or her own right. Similar compliance regulations are

not applicable within the government sector.

Partial eta squared values of .028 and .085 indicated small effect sizes. However,

medium effect sizes of d = .476 were calculated between employees working in the

financial sector and those in the government sector with regard to the Responsibility

dimension and d = .423 between employees employed in the financial sector and those

employed in other sectors. With regard to the Territoriality dimension, the following

results were found: a relatively high medium effect size of d = .753 between employees

in the financial and government sector, between employees in the financial and other

sectors (d = .735) and between employees in the chemical industry and government

sector (d = .785). Medium effect sizes were also calculated between those employed in

the chemical industry and other sectors (d = .446) and those working in the government

and manufacturing and production sectors (d = .546).



Table 4.29: ANOVA and post-hoc Scheffés test – Sector in which the organisation operates

Dependent variables


Partial eta squared

Identification 1.456 .214 Financial services

103 70.592 14.803 .008

Chemical / Petroleum

186 71.839 12.143

Government 172 72.523 13.102 Manufacturing

& production 110 74.691 11.846

Other 123 Responsibility 4.880 .001* Financial

services 103 42.67ab 3.725 .028


186 41.634 3.243

Government 172 40.506a 4.953 Manufacturing

& production 110 41.164 4.374

Other 123 40.919b 4.479 Autonomy .488 .744 Financial

services 103 28.301 5.490 .003


186 27.538 5.476

Government 172 27.680 5.102 Manufacturing

& production 110 28.046 5.007

Other 123 27.553 5.330 Territoriality 16.084 .000* Financial

services 103 16.301ab 4.907 .085


186 16.204cd 4.766

Government 172 19.913ace 4.658 Manufacturing

& production 110 17.182e 5.297

Other 123 18.626bd 5.470 *p < .05

• Level in the organisation

The four different levels on which people operate in their organisations were compared.

The different levels are as follows: operational level, junior management level, middle

management level and senior management level. From the ANOVA results displayed in



Table 4.30, it is evident that significant (p < .05) differences exist between the four

groups with regard to the different dimensions that psychological ownership comprises.

Post-hoc comparisons done by means of the Scheffé test reveal significant differences

among the four groups in the following dimensions:

• Identification

Once again significant differences (F = 5.392; p < .05) exist between employees

functioning on operational level and junior management level and those on senior

management level. Employees on senior management level (mean = 76.450) showed a

higher sense of identification toward their organisations than those employees

functioning on operational level (mean = 70.873) and junior management level (mean =

71.240). Senior management may feel a stronger sense of ownership and feel “part of

the pride” because normally they have been in the organisation for quite some time.

These people are usually heads of departments in which they play a dominant role and

where they take decisions on both human and asset capital and see the effect of the

decisions within the organisation. This may enhance their feelings of attachment to the

organisation. They feel they have earned their wings”. Govender and Parumasur (2010)

for example, found that top managers possess and display significantly higher levels of

director competencies than middle and senior managers. (Competencies in the director

role entail communicating a vision, setting goals and objectives, and designing and

organising.) The partial eta squared value calculated was small (.023). Cohen’s d

(1992) was calculated between employees on operational level and senior management

level and a medium effect size of (d = .434) was revealed. A medium practical

significance (d = .414) was found between junior management and senior management

level employees.



• Responsibility

There is a significant difference (F = 5.483; p < .05) between employees on operational

level and junior management level and those employees on senior management level

with regard to Responsibility. Employees on senior management (mean = 42.450) level

showed a slightly higher inclination towards responsibility in their organisations than

those on operational level (mean = 40.910) and junior management level (mean =

40.653). Both socially and by law, there is strenuous responsibility placed on senior

management and they can be held liable for major disasters such as health and safety

breaches, and can be personally fined or receive a jail sentence. According to Wood

and Winston (2007, p.169), leaders should be accountable, therefore need to accept the

responsibilities inherent in their leadership position, “not just for activities,

circumstances, or past results, but for future direction, potential effectiveness, possibility

thinking, an inspiring shared vision, and maximum contribution”. The partial eta squared

value calculated was small (.023). Calculation of the effect sizes revealed a medium

practical significance (d = .331) between employees on operational level and those on

senior management level. A medium effect size value of d = .422 was also found

between employees functioning on junior management level and those on senior

management level.

• Autonomy

With regard to the Autonomy dimension, a significant difference (F = 10.094; p < .05)

exists between employees on an operational level (mean = 26.864) and those on middle

management level (mean = 28.448) and senior management level (mean = 29.592).

Both employees on middle management level and senior management level experience

more autonomy in the workplace than those who function on an operational level. A

significant difference also exists between employees on junior management level (mean

= 26.813) and those on middle management level (mean = 28.448) and senior

management level (mean = 29.592). Both employees on middle management level and

senior management level experience more autonomy in the workplace than those who



function on junior management level. The reason for this difference probably lies in the

fact that the higher up the employees are in the organisation, the broader their mandate

to make decisions and the bigger the impact of those decisions will be on the long-term

sustainability of the organisation. This is supported by Brass (1985), who observed that

employees exposed to high job design autonomy experienced more influence (control)

than their counterparts working with low autonomy. A small partial eta squared value of

.042 was calculated for the autonomy dimension. Medium practical significance was

reflected between operational and middle management level employees (d = .308),

between operational and senior management level employees (d = .516), between

junior management and middle management employees (d = .326) and between junior

management and senior management employees (d=.538).

• Territoriality

The following groups differ significantly (F = 5.284; p < .05) with regard to the

Territoriality dimension: employees functioning on the operational level (mean = 18.733)

showed a higher territorial need than employees functioning on senior management

level (mean = 16.575). Operational staff displays an attachment to their

department/organisation for the purposes of constructing, communicating, maintaining,

and restoring (Brown et al., 2005). This is done to ensure their department access and

to maintain their existing capacity and resources despite the general cut-back

experienced by many organisations in the current economic downturn. Further

indication of territorial behaviour on operational level is that operational teams mark

their areas with personalised insignia (Brown et al., 2005). Senior management showed

a lower territorial need, as many of them have their own office to which they feel

attached, while operational staff normally works in open-plan offices where personal

space is limited. Although a partial eta squared value for the Territoriality dimension

calculated was small, a medium effect size of d = .407 was reflected between

operational and senior management level employees.



Table 4.30: ANOVA and post-hoc Scheffé’s test – Level in the organisation

Dependent variables

F – value Sig. Sub-groups N Mean Standard deviation

Partial eta squared

Identification 5.392 .001* Operational 221 70.873 a 13.375 .023 Junior

management 150 71.240 b 12.853

Middle management

201 72.751 12.925

Senior management

120 76.450 ab 12.218

Responsibility 5.483 .001* Operational 221 40.910 a 4.495 .023 Junior

management 150 40.653 b 3.629

Middle management

201 41.647 3.547

Senior management

120 42.450 ab 4.784

Autonomy 10.094 .000* Operational 221 26.864 a b 5.663 .042 Junior

management 150 26.813cd 5.422

Middle management

201 28.448 ac 4.547

Senior management

120 29.592 bd 4.857

Territoriality 5.284 .001* Operational 221 18.733 a 5.089 .023 Junior

management 150 17.260 5.113

Middle management

201 17.617 4.976

Senior management

120 16.575 a 5.497

*p < .05

• Number of years working in the current organisation

However, contrary to general belief, Table 4.31 indicates that no significant differences

exist between the number of years employees have worked in their current organisation

and the dependent variables. Pierce et al. (2001) propose that a longer association with

a target (e.g. long tenure) will probably lead to perceptions of knowing the target better

and as a result, to a sense of ownership.



Table 4.31: ANOVA and post-hoc Scheffé’s test – Number of years working in the current organisation

Dependent variables

F - value Sig. Sub-groups N Mean Standard deviation

Partial eta

squared

Identification 3.709 .464 Less than 5 years

307 70.981 14.013 .015

6-10 years 138 72.044 12.667 11-20 years 163 74.282 11.341 21+ years 105 74.895 12.187 Responsibility 1.449 .227 Less than 5

years 307 41.283 4.087 .006

6-10 years 138 41.261 4.338 11-20 years 163 40.761 4.722 21+ years 105 41.857 3.707 Autonomy 1.718 .162 Less than 5

years 307 27.365 5.398 .007

6-10 years 138 27.790 5.306 11-20 years 163 27.988 4.765 21+ years 105 28.657 5.288 Territoriality .372 .774 Less than 5

years 307 17.808 4.968 .002

6-10 years 138 17.768 4.972 11-20 years 163 18.147 5.507 21+ years 105 17.476 5.693 *p < .05

• Number of years working in current position

According to the results displayed in Table 4.32, significant differences exist between

the numbers of years employees have worked in their current position in the

organisation with regard to the Territoriality dimension (F = 6.638; p < .05). Once more a

Scheffé post-hoc test was conducted to determine where the differences lie between

the groups. The results indicated that employees employed for less than 5 years in their

current position differ significantly from employees who have worked between 6 and 10

years and 11 to 20 years in their current position in terms of their territorial need.

Employees who have worked between 6 and 10 years (mean = 18.971) and 11 and 20

years (mean = 19.131) in their current position indicated a higher territorial need than

employees employed for less than 5 years (mean = 17.214) in their current position.



Employees working in their current position for 11 to 20 years have reached corporate

maturity and have made their position absolutely their own. This is good news for the

organisation, which has employees that are well entrenched in their position and form

the backbone of the organisation. The downside of this is that these employees may

have fallen into a comfort zone and may no longer grow with the organisation. They

display a higher territorial score since they are protecting their current environment and

will oppose any change in their operating environment. Prevention-orientated people

tend to be less open to change and they prefer to stick with the already known (Kark &

Van Dijk, 2007). It seems that employees who have worked in their current position for

11 to 20 years are more prevention-orientated and therefore more territorial than

employees employed for less than 5 years in their current position.

A small partial eta squared value of .027 was calculated. However, medium effect sizes

were calculated between employees employed for less than 5 years in their current

position and those employed between 6 and 10 years (d = 0354) as well as between

employees working in their current position for less than 5 years and those employees

working between 11 and 20 years (d = .356).



Table 4.32: ANOVA and post-hoc Scheffé’s test – Number of years working in current position

Dependent variables


Partial eta squared

Identification 1.935 .123 <5 years 471 72.064 13.516 .015 6-10 years 102 73.431 11.818 11-20 years 99 71.950 12.602 21+ years 40 76.829 9.633 Responsibility .834 .475 <5 years 471 41.420 4.200 .006 6-10 years 102 40.912 4.315 11-20 years 99 40.818 4.687 21+ years 40 41.073 3.357 Autonomy .775 .508 <5 years 471 27.796 5.095 .007 6-10 years 102 27.480 5.734 11-20 years 99 27.556 5.397 21+ years 40 28.878 5.124 Territoriality 6.638 .000* <5 years 471 17.214ab 5.077 .027 6-10 years 102 18.971a 4.839 11-20 years 99 19.131b 5.665 21+ years 40 18.902 5.253 *p < .05

4.3 CONCLUSION

In this chapter a measure of psychological ownership was developed by following eight

steps for scale development as suggested by Hinkin (1998), DeVellis (2003), and

Spector (1992).

In the first step all the dimensions underlying the concept of psychological ownership

were clearly defined and described, namely self-efficacy, self-identity, sense of

belonging, accountability, autonomy, responsibility and territoriality.

In the second step, by following a deductive approach, items were generated for each

one of the dimensions. A total number of 54 items were generated.

In the third step of scale development a Likert-type rating scale, with an equal 1-6

agreement format, was chosen.



In the fourth step a panel of nine subject matter experts reviewed the initial pool of items

to judge each item related to the specific dimension of psychological ownership. The

experts commented on the item content, item style, and comprehensiveness of the

instrument. After the application of Lawshe’s (1975) quasi-quantitative technique a total

number of 34 items were retained. However, a total of 11 items were kept despite

Lawshe’s application. Based on the suggestion of the subject matter experts, an

additional 24 items derived from the literature study were added to the measurement to

better represent the total content domain. The total number of items to be included in

the final instrument to be tested was 69. Preliminary analysis on the pilot study (n = 46)

revealed satisfactory reliability coefficients for each of the dimensions.

In the fifth step the instrument was administered to a non-probability convenience

sample (N = 712) comprising employed professional, high-skilled and skilled individuals

in various organisations in both the private and public sector in South Africa. However,

the sample was randomly split into two subsets. A sample of 356 was used for the

development of a model and the remaining half was used for validating the results that

were attained from the first half (Anderson & Gerbing, 1988).

In the sixth step of scale development an exploratory factor analysis (EFA) was

performed on the sample comprising 356 individuals. The results of the Bartlett’s Test of

Sphericity and the KMO measure of sample adequacy confirmed that the data was

suitable for factor analysis. The parallel analysis signified four significant factors. Two

rounds of exploratory factor analysis were performed on the four-factor model that

resulted in a measure comprising 35 items. The four factors of the South African

Psychological Ownership Questionnaire (SAPOS) were labelled Identification,

Responsibility, Autonomy and Territoriality respectively. The four factors retained

explained 50.7% of the total variance of the data set.

Results of the second-order factor analysis clearly indicated the existence of two

distinctive dimensions, labelled promotive (promotion-orientated) psychological

ownership and preventative (prevention-orientated) psychological ownership



respectively. The Tucker’s phi coefficients for the Black and White respondents were all

acceptable (>.90) indicating that the four factors of the SAPOS were equivalent for the

two race groups. All four subscales revealed highly satisfactory Cronbach alpha

coefficients (0.94; 0.87; 0.87 and 0.78). Results of the descriptive statistics for the four

scales of the SAPOS indicated a deviation from the normal distribution with a tendency

towards negative skewness and leptokuric distributions.

In the seventh step the four-factor model consisting of 35 items of the SAPOS was

subjected to confirmatory factor analysis on the other half of the sample (n = 365).

Mardia’s coefficient indicated that the data was non-normally distributed and therefore

the robust maximum-likelihood (ML) estimation with the Santorra-Bentler scaled chi-

square was employed. Except for the NNFI value of 0.897, the chi-square/df ratio (1:7),

CFI (0.904), RMSEA (0.045), and SRMR (0.059) values met the minimum

recommended standards, indicating a reasonable fit.

In the eighth step the discriminant validity and criterion-related validity of the SAPOS

were examined. Examination of the variance extracted estimates confirmed discriminant

validity within the model. The relationship between psychological ownership and similar

constructs was examined to develop a semantic network. Promotive psychological

ownership was positively correlated with affective commitment and job satisfaction and

negatively related to turnover intentions. These results provided evidence of criterion-

related validity.

To determine the differences between the means scores of the different groups with

regard to their biographical characteristics, independent sample t-tests and analysis of

variance (ANOVA) were conducted. Differences exist between groups with regard to

their ethnic grouping, gender, registration with a professional board, age, educational

level, sector in which their organisation operates, level in the organisation and number

of years working in their current position. However, in most cases medium practical

significance was established between the groups. Contrary to general belief, no



significant differences were found between the number of years employees had been

working in their current organisation and the dependent variables.

The following chapter will conclude the research and discuss a summary of the study. It

will assess the contribution of the study from a theoretical, methodological and practical

point of view. The limitations and directions for future research will be discussed.

CHAPTER 4 RESEARCH RESULTS AND FINDINGS

Documents

Transcript of CHAPTER 4 RESEARCH RESULTS AND FINDINGS